State-of-the-Art Review of Current Technology in Pressure Injury Early Detection

Abstract

OBJECTIVE:

To review currently available devices on pressure injury (PI) early detection, summarize challenges and opportunities to clinical implementation, and propose evaluation standards for device categories.

DATA SOURCES:

PubMed and US Food and Drug Administration (FDA) databases.

STUDY SELECTION:

Published in English from peer-reviewed journals with full text available. Excluded if opinion statements, lack of empirical data, or unrelated to project’s objective.

DATA EXTRACTION:

For both clinical device and research equipment: measurement mechanisms, measurement types, outcome/output, FDA classification, and indications for use. Addition data were extracted for clinical devices: instruction for use, end user, order requirement, and billable code.

DATA SYNTHESIS:

The 4 clinical devices are ultrasound, long-wave infrared thermography, subepidermal moisture assessment, and nearinfrared spectroscopy. The 3 research devices are laser Doppler flowmetry, laser speckle contrast imaging, and colorimetry.

CONCLUSIONS:

The measurement mechanisms of all devices are unique and different from each other. One commonality is that they could measure the nonvisual signs of PI (eg, inflammation, edema, ischemia, and hypoxia) except colorimeter. Some clinical devices are promising to assist with early identification of PIs, especially in individuals with dark skin tones. Currently, there is no reimbursement available for early detection of PI. Current evidence did not support replacing the standard skin assessment of visual inspection and palpation with the devices reviewed, rather using validated devices to augment the current practice standard. This is especially recommended for individuals identified as high risk for a PI on admission to a facility.

INTRODUCTION

Pressure injury (PI) early detection is a constant surveillance process that requires a systematic and diligent approach. It is part of the prevention and treatment strategies continuum whereby clinicians seek to avoid PI formation (prevention) or thwart further deterioration (treatment) of a PI. The current standard of assessing for a PI requires a risk and skin assessment. Risk assessment process includes using a validated instrument as well as accounting for other intrinsic and extrinsic risk factors not in risk assessment instruments.¹ Skin assessment consists of visual and tactile measures to detect the presence of a PI.²

Clinicians are always seeking different methods and indicators to augment the current assessment standard detect a PI early. Two recent recommended methods have included the use of a PI early detection device and skin tone assessment guides. The early detection devices are sought as a means to “visualize” possible damage below the skin level that cannot be seen by the naked eye.^3,4 Skin tone assessment guides are based on the concentration of epidermal melanin, which is classified in 2 types—darker eumelanin and lighter pheomelanin, and is responsible for the prime determinants of skin color.⁵ The primary reason for employing skin tone assessment guides is that certain colors and hues exhibited by damaged tissue, such as purples, maroon, and deep blues, can be masked by a dark skin tone color and therefore can lead to a missed identification opportunity in early PI dectection.⁶

Devices developed for early detection identify nonvisible changes in tissue characteristics that may be precursors to PI, that is, changes in characteristics that occur before any visible signs of tissue damage to naked eyes or changes in characteristics that are difficult to assess consistently through visual inspection alone. This information can assist in clinical decision-making by implementing strategies to prevent further tissue deterioration and subsequent PI development. The goal of this review was to compare the devices available to provide PI early detection and summarize (1) interdevice variations in the mechanisms of measurement, (2) considerations when selecting a device, (3) insurance coding status, and (4) the strategies to develop standards that evaluate important operating parameters and device characteristics.

METHODS

A team from the National Pressure Injury Advisory Panel, convened for purposes of this project, included clinicians, researchers, and Corporate Advisory Council members, who were knowledgeable and had various experiences with PIs. To generate a list of devices for this review, one author conducted a literature search through PubMed to generate a list of publications regarding PI early detection devices using search terms: “pressure injury,” “pressure ulcer,” “HAPI,” “HAPIs,” “Deep Tissue Pressure Injury,” “suspected deep tissue pressure injuries,” “imaging,” “infrared,” “Doppler,” “Spectrometry,” “moisture scan,” “SEM,” “subepidermal scan,” “subepidermal scan,” “Impedance,” “bioimpedance,” “ultrasound,” ‘ultrasonograph,” “Thermography,” “Long-Wave Infrared thermography,” and “Temperature.” The team reviewed the list of publications and finalized a master list of devices for review based on expert consensus. Devices were then categorized either as clinical devices or research equipment. The clinical device list was determined based on criteria: (1) available in the US at the time of the team’s review and synthesis of data, (2) received US Food and Drug Administration (FDA) authorization or clearance, and (3) had an indication for health care professional use. The research device list was determined based on criteria: (1) available in the US at the time of review, (2) received FDA exempt, and (3) had been utilized in human research studies for nonvisible tissue characteristics of PIs.

A comprehensive literature search for each device category was performed independently by up to 2 contributing authors to identify research studies, review articles, meta-analyses and clinical trials related to early PI identification. Publication search filters included articles that were: (1) in English, (2) from peer-reviewed journals, and (3) available in full text. Publications were excluded if opinion statements, lack of empirical data, or unrelated to project’s purpose. The FDA 510(k) Premarket Notification Database and the FDA De Novo Classification Database were also queried to gather information on device classification, approved indications for use, safety profiles and other pertinent regulatory information. In addition, manufacturer websites, medical device directories, and market reports were searched for commercially available devices. Technologies were excluded if they were discontinued by the manufacturer.

The following data were extracted for comparisons among devices within clinical devices and research equipment separately: (1) measurement mechanism (nonvisible tissue characteristics, eg, temperature, edema, blood flow), (2) measurement types (quantitative/qualitative), (3) outcome/output (ie, results of the measurement, eg, image, number, percentage), (4) FDA classification, and (5) indications for use. Addition data were extracted for clinical devices for comparison: (1) instruction for use, (2) end user, (3) order requirement and (4) billable code.

RESULTS

Clinical Devices

Tables 1 and 2 provide the list of clinical devices under 4 categories: ultrasonography, long-wave infrared thermography (LWIT), subepidermal moisture measurement (SEM), and near-infrared spectroscopy (NIRS).

TABLE 1.

FDA DEVICE CLASSIFICATION AND INTENDED USE SUMMARY FOR MARKET-APPROVED/AUTHORIZED MEDICAL DEVICES

Technology	FDA Classification	Regulatory Class	Regulation Name	21 CFR* Regulation Number	Product Code	Indication for Use	Measurement Mechanism	Measurement Type	Outcome/Output	Available Countries
Ultrasonography	510(k)	Class II	Radiology	892.156	LWI, IYO, NQQ†	Project a pulsed sound beam into body tissue to determine the depth or location of the tissue interfaces and to measure the duration of an acoustic pulse from the transmitter to the tissue interface and back to the receiver	Changes in structural tissue properties: elasticity, stiffness, unclear layered structure, hypoechoic lesions, discontinuous fascia and heterogeneous hypoechoic areas	Qualitative	Imaging and radiology reports	Globally
LWIT thermography	510(k)	Class I	Surgical camera and accessories	878.416	FXN	An adjunctive tool to help a trained and qualified health care professional measure and record external wound and body surface data	Infrared-based thermal intensity	Quantitative	Photographs and thermal infrared images	US
SEM measurement	De Novo	Class I	Pressure ulcer management tool	876.21	QEF	An adjunct to standard of care when assessing the heels and sacrum of patients who are at increased risk for PI	SEM or localized edema	Quantitative	Measurements of localized edema as a comparison of multiple SEM values across the anatomy displayed as a SEM ∆	US, UK, and 28 other countries
Near-infrared spectroscopy	510(k)	Class II	Oximeter, tissue saturation	870.27	MUD	To determine oxygenation levels in superficial tissues	Soft tissue oxygenation/oxygen saturation	Qualitative	Displays a two-dimensional, color-coded image of the tissue oxygenation	Canada, US

^* CFR: Code for Regulations; LWI: ultrasound, sinus; IYO: system, imaging, pulsed echo, ultrasonic; NQQ: system, imaging, optical coherence tomography; FXN: tape, camera, surgical; QEF: pressure ulcer management tool; MUD: oximeter, tissue saturation;

^† Multiple FDA product codes exist for predicate or substantially equivalent ultrasonography devices available through multiple manufacturers.

Abbreviations: LWIT, long-wave infrared thermography; PI, pressure injury; SEM, subepidermal moisture.

TABLE 2.

DEVICE CHARACTERISTICS

Technology	Principle of Operation	Device User	Body Surface Contact	Require Order	CPT Code	ICD-10-PCS Code	Settings for Use
Ultrasonography	Diagnostic ultrasound imaging enabling visualization and measurement tool. An investigation of choice when the organ (skin tissue) of clinical concern is accessible and assessable for sound wave interrogation	Radiation Specialists	Contact, noninvasive	Yes	None	B54___	Inpatient, outpatient
LWIT thermography	Capture thermal images through infrared radiation emitted by the human body to measure the thermal intensity data of a part of the body or 2 body surfaces	Health care professionals	Noncontact, noninvasive	NA	None	None	Inpatient, outpatient
SEM measurement	Detects localized edema by comparing biocapacitance of subdermal tissues. The greater the difference between the region of interest and adjacent tissue, the greater the localized edema at the specific anatomy	Health care professionals	Contact, noninvasive	NA	None	XX2KXP9	Inpatient, outpatient
Near-infrared spectroscopy	Performs spectral analysis at each point in a two-dimensional scanned area to determine the approximate values of oxygen saturation (StO2), oxyhemoglobin levels (Hb02), and deoxyhemoglobin levels (Hb) in superficial tissues	Health care professionals	Noncontact, noninvasive	NA	0640T, 0859T, 0860T	8E02XDZ	Inpatient, outpatient

Abbreviations: LWIT, long-wave infrared thermography; SEM, subepidermal moisture.

Overview of the Technologies

Ultrasonography (excluding point-of-care ultrasonography) is a diagnostic imaging technology relying on the emission and measurement of ultrasonic sound waves emitted over human tissue. High frequency (>20MHz)⁷ ultrasound was the most commonly used in previous PI studies, which can provide anatomic images and mechanical properties (ie, softness/hardness) of tissues. Ultrasound could detect pockets of fluid/edema at different tissue layers, which could serve as early signs of PI.⁸ In addition, it could be used for early detection of deep tissue PIs (DTPI) by measuring changes in tissue structures underneath intact skin, such as unclear layered structures, hypoechoic lesions, discontinuous fascia and heterogeneous hypoechoic areas.⁹

The LWIT is an imaging technology that measures the body heat produced by tissue metabolism and blood distribution, which reflects the presence or absence of perfusion of dermal and subcutaneous tissues under intact skin or open wound.¹⁰ The temperature pattern provides an objective measurement of the tissue that can be identified by an increased temperature (eg, inflammation) or decreased temperature (eg, ischemia or tissue death) when compared with adjacent area.^10,11 LWIT could be used for early PI detection^3,12–15 by quantifying the temperature changes at the pressure area as warmer or cooler to adjacent tissue.^14,16–19

Subepidermal moisture (SEM) assessment technology utilizes biocapacitance sensors to detect early-stage PI by measuring localized edema underneath the skin at different anatomic locations.²⁰ This localized edema or persistent focal edema is one of the earliest detectable indicators of tissue damage resulting from pressure-induced inflammatory responses following tissue deformation.²¹ The sensors measure variations in the electrical capacitance of tissue, which directly relate to changes in SEM levels caused by changes in localized edema, providing objective evidence of early-stage pressure injury development before visible damage occurs.

Multiple SEM measurements are made across a specific anatomy. The greater the difference between SEM measurements at and around bony prominences of the anatomy of interest, the greater the localized edema at the specific anatomy.²⁰

The NIRS is an imaging technology that measures tissue oxygenation by penetrating near infrared (wavelength 750 to 2500 nm) and collecting the reflectance wavelength range (600 to 1000 nm) corresponding to oxygenated hemoglobin, deoxygenated hemoglobin and melanin. Tissue oxygen saturation (StO₂) is the fraction of oxygenated hemoglobin to total hemoglobin. Changes in StO₂ could reflect tissue viability and serve as early signs of PI. This device could be applied to wound care, limb preservation and reconstructive surgery.²²

Similarities and Differences Across Devices

The commonality for the clinical devices is noninvasive measurements with the ability to identify a specific early PI structural or functional change(s) (eg, edema, inflammation, ischemia, hypoxia) not visible to the human eye. Given that these devices provide different forms of measurement, the FDA regulatory classes and names vary.

Indication for Clinical Use

Ultrasonography is indicated for use as a diagnostic tool. This includes assessment for diagnostic purposes or using ultrasound guidance for interventional procedures, such as therapeutic injections or draining an anatomic area. It is the device of choice when an organ of clinical concern is accessible for sound wave interrogation. Results are best when the localized area of concern is closer to the ultrasound transducer.

The LWIT device is indicated for capturing visual and thermal images; the visual to measure the diameter, surface area, and perimeter of a part of the body or 2 body surfaces, and the thermal image to measure the temperature. Both components of the device (ie, visual and thermography) are noncontact with respect to the patient and provide an adjunctive tool to help a trained and qualified health care professional measure and record any temperature variation and the size measurements of open wounds and areas of nonvisual changes in intact skin to identify the early changes underneath.

The SEM ∆ device is a contact-based, hand-held point-of-care diagnostic device indicated for measuring subepidermal moisture or localized edema underneath the skin. Comparisons of multiple measurements across the anatomy identifies the maximum difference in values, displayed as a “SEM-∆.” Increased SEM ∆ indicates increased localized edema following the inflammatory process resulting from deformation, and the manufacturer recommends intervention when a SEM ∆ is ≥0.6. Daily SEM ∆ assessments are intended to be used by health care professionals as an adjunct to the standard of care when assessing the heels and sacrum of patients who are at increased risk for PIs.

The NIRS is intended for use by health care professionals as a noninvasive tissue oxygenation measurement system that reports an approximate value of StO₂, relative oxyhemoglobin and deoxyhemoglobin levels.

Scope of Practice

Devices currently authorized or cleared by the FDA for commercial use (Table 1) include a LWIT for thermal imaging, SEM assessment technology as a PI management tool to measure and monitor localized edema or persistent focal edema in patients at risk of developing PIs, and the NIRS for oximetry and tissue saturation. Each of these hand-held, point-of-care, noninvasive devices provide unique benefits and implications for deployment in clinical practice. Device training varies between different technologies and largely depends on its intended use and the end user (eg, clinician or other health care provider). LWIT and NIRS devices are noncontact but require bedside clinicians to train on thermal and spectral image analysis and interpretation, whereas SEM assessments require users to scan across the anatomy at risk using single-use sensors and requires light contact. Further, differences exist in the clinical interpretation of measured outcomes between different technologies. Thermography devices suggest inflammation and ischemia depending on the relative increase or decrease in thermal intensity respectively determined by the relative temperature of an adjacent control measurement site.¹⁹ Subepidermal moisture assessments with a value ∆≥0.6 indicate localized edema or developing damage on tissues, but with intact skin only.²³ Similarly, decreased StO₂ from NIRS describe compromised perfusion and circulation in vascular structures.

Billable Codes for Insurance

General considerations when assessing if a code will reimburse its assigned dollar amount of payment are to: (1) determine if the code specifically states the intended use (eg, for PI detection, wound assessment, flap assessment or other), as an audit may result in repayment if the Current Procedural Terminology (CPT) codes are not used for the specific description given, (2) understand the type of CPT code (ie, category I, or typical codes, or category III, or temporary CPT codes for new and developing technology),²⁴ (3) know if the CPT code and Revenue codes are compatible with the place of service the device is being used, and (4) document the procedure, outcome and medical necessity (billable codes require supporting documentation).

Table 2 contains billable code of the clinical devices included in this review. Only ultrasonography has billable codes but no CPT codes for PI detection. NIRS has class III codes, which the payment amount may not be reimbursed by all payers or may be reimbursed in one state but not others based on the insurance provider.²⁴ The LWIT does not have CPT codes for billing. The SEM has a new technology procedure code for the measurement and monitoring of localized edema (ICD-10-PCS XX2KXP9).

Nonvisible changes in tissue characteristics (changes that occur before any visible signs of tissue damage to naked eyes or changes that are difficult to assess consistently through visual inspection alone) provided by these clinical devices may assist with clinical decision making, such as tailored prevention interventions to mitigate or decrease PI severity and hospital-acquired PIs (HAPI). Decreased HAPI can deliver a financial return on investment versus looking only at a financial benefit from procedure billing.

Psychometric Properties of the Devices in PI Prediction

Because the application of these devices on PI detection is not yet fully studied, the psychometric properties in PI detection reported in the literature are limited to certain stage and location of PI and patient population. Ultrasound was reported to have sensitivity of 100% and specificity of 74.8% to 84.7% for DTPI in emergency department patients at high risk of PIs (n = 33). The overall accuracy for detecting DTPI was reported to be 84.9%.²⁵

SEM assessment technology reported high inter-operator and inter-device reliability with the intraclass correlation coefficient (ICC) exceeding 0.80 (95% CI). The device demonstrates a sensitivity of 82% to 100% and specificity of 32.9% to 83% for Stage 1 and/or DTPI in acute and postacute patients for the sacrum and heels. Reported areas under the curve (AUC) of 0.7809 to 0.9181 (95% CI, P < .0001) significantly exceeded clinical judgment alone in the early detection of PI.^26–28

LWIT was reported to have low within-reader variation (1%) and between-reader variation of mean temperature (2%) during initial patient encounter. During follow-up patient encounter, the LWIT was reported to have low between-reader variation (2%).¹⁰ This suggested a high agreement within and between readers. Although the sensitivity and specificity data on PI prediction are yet to be examined, one study reported that LWIT demonstrated zero false-negative of hospital-acquired DTPI.¹⁵

Considerations When Implementing a PI Early Detection Device

The rationale for opting to use early PI detection devices is multifactorial. These factors should be discussed among staff with different roles (eg, clinical, finance, supply chain, information technology, education) commencing before a device is purchased or rented.

Clinical implementation factors to be considered would include professional qualifications and training of staff utilizing the device, availability of staff to perform the assessments (ie, 24/7, target patient population(s), etc), and infection control recommendations. Other factors include device specific attributes, information technology concerns, and financial implications such as: (1) the target patient population, (2) the device’s indications for use, accuracy (sensitivity, specificity), (3) scope of practice of the operator(s), (4) troubleshooting, biomedical and technical support from the manufacturer, (5) ease of initial training and ongoing support for staff, (6) electronic health record (EHR) compatibility and data security, and (7) billable service and cost/benefit of the technology tied to outcomes.

Choosing the appropriate device is based on the patient population, care facility, number of devices required versus the number of end-users, and the end-user’s scope of practice. For example, highly trained individuals in the field of radiology are required to perform diagnostic ultrasound measurements, while qualified health care professionals are required for the remaining 3 devices. In addition, because ultrasonography is heavily dependent on radiologic infrastructure including machines and licensed/certified personnel, especially on results interpretation, integrating ultrasound without such infrastructure may not be practical for settings such as skilled nursing facilities and long-term care. It is recommended to consult with other similar clinical settings before deciding to invest in any technology.

The financial consideration when planning to introduce new technology involves determining costs associated with: (1) hardware, (2) licenses, (3) disposables and supplies, (4) personnel, (5) education and training, (6) reimbursement potential when the device(s) is utilized in practice, (7) cost of information technology (IT) integration of the device into the EHR, (8) facilities’ IT infrastructure, and (9) cybersecurity analysis of some devices. In addition to determining the costs incurred when integrating a new technology, the benefits realized from daily use are equally important. Since these technologies are considered for use in advanced clinical assessment, rather than for diagnosis, one method for evaluating cost may be to assess cost avoidance of incurring a PI. This is a viable metric because early PI detection will avoid costs incurred for treating a more severe PI that can include dressing materials, surgery, imaging, laboratory tests, nursing care and an increased hospital stay.^29,30 Having conversations with other similar organizations using a technology may provide insight to other metrics worth assessing.

Research Equipment

Tables 3 and 4 contain the list of research equipment for early PI detection under 3 technological categories: laser Doppler flowmetry (LDF), laser speckle contrast imaging (LSCI), and colorimeter.

TABLE 3.

FDA DATA OF THE RESEARCH DEVICE CATEGORIES

Technology	FDA Status	FDA Device Area	CFR	FDA Product Code	FDA Classification	Available Countries
Laser Doppler flowmetry system	510(k)	Cardiovascular	870.2120/870.2100	DPT/DPW	Class II	Globally
Laser speckle contrast imaging	510(k)	Cardiovascular	870.2120	DPT	Class II	Globally
Colorimeter	510(k)	Clinical chemistry and clinical toxicology devices	862.2300	JJQ	Class I	Globally

DPT: probe, blood-flow, extravascular; DPW: flowmeter, blood, cardiovascular; JJQ: colorimeter, photometer, spectrophotometer for clinical use.

TABLE 4.

MEASUREMENT, OUTCOMES, INSTRUCTION FOR USE AND LIMITATIONS FOR THE RESEARCH DEVICE CATEGORIES

Technology	Measurement Mechanism	Measurement Type (Quantitative/Qualitative)	Outcome/Output	Instruction for Use	Limitations
Laser Doppler flowmetry system	Measure blood flow by projecting a laser beam to the skin and collect the scattered light through the optical fibers	Quantitative	Blood flow in arbitrary units or absolute flow units such as mL/min/100 g of tissue. Variables could be derived from hyperemic responses and compared within or between individuals	Place on the skin directly for measurement. For comparison within or between individuals, the measurement probe should be placed at the same anatomic location (eg, bony landmark as reference)	Due to the nature of the blood flow in capillaries, researchers and manufacturers generally agree that it is not appropriate to use absolute flow units Due to its arbitrary unit, comparison between measurements and different devices should be interpreted with caution
Laser speckle contrast imaging	Blood perfusion imager based on laser speckle contrast analysis technology	Quantitative	Color images ranging from blue to red showing perfusion from low to high	Color camera of perfusion of a larger area (eg, whole hand) rather than one single point (laser Doppler). Does not require contact to the skin. Provides color map of perfusion distribution in the region of interest	Measurement depth varies between machines and may not be comparable between different machines and different people
Colorimeter	Measurement of red, green, and blue light reflected from the skin to calculate descriptive parameters of skin color in a variety of color notation systems	Quantitative	L*a*b* color notation. L*: “luminance” or “brightness” of a color, 0=black, 100=white a*: a color falls on a continuum from red (+) to green (-) b*: a color falls on a continuum from yellow (+) to blue (-) Some devices reports melanin and erythema indices based on the L*a*b measurements	The device illuminates the skin and photodetectors collect reflected light in the red, green, and blue regions of the visible light spectrum (∼600 nm, 560 nm, and 450 nm, respectively)	Requires skin contact Erythema and melanin metrics are typically proprietary. Calculations are not generally reported by manufacturers and are not directly comparable across devices

Overview of the Technologies

The LDF is a noninvasive device that measures skin blood flow based on Doppler shift. The laser light penetrates the skin, and the optic probe captures the laser light reflected when hitting moving red blood cells. By quantifying the frequency shift of the scatter light, the velocity of the moving red blood cells could be measured, and subsequently, the amount of blood flow through the area of interest (up to 2 mm beneath the surface of the skin) could be quantified. This device has been used in the field of physiology to quantify microvascular function such as ischemia (reactive hyperemia) and localized heating (heat hyperemia).³¹ Traditionally, applications of LDF in PI research were mainly conducted to compare differences in skin blood flow under mechanical load between people with increased risk of developing PI versus controls,^32,33 or to compare differences in skin blood flow on different support surfaces, and to evaluate pressure redistribution strategies.^34–36 By quantifying the changes in skin blood flow, or changes in the microvascular function, this device may serve as a method to detect early signs of PI, such as a decrease in perfusion or altered microvascular function. One proof-of-concept study showed that the heat hyperemia was diminished in acute and subacute hospitalized patients with spinal cord injury before PI development before discharge from the hospital, suggesting that an altered microvascular function could serve as a nonvisible sign of PI.³⁷

The LSCI is a noninvasive imaging device that captures moving object patterns, such as blood flow, utilizing laser speckle contrast analysis.³⁸ The images are obtained by diffusing laser light on the region of interest, generating speckle pattern of the area and capturing the speckle pattern with a camera. Postprocessing software then transfers the speckle pattern into speckle contrast values and demonstrates the “flow” of that area through speckle contrast imaging on the computer screen. This device has been used to visualize skin blood flow for predicting burn wound healing with good validity³⁹ and predicting venous ulcer healing with good sensitivity and specificity.⁴⁰ By quantifying the changes in skin blood flow of the target area and monitoring the blood flow in real time, this device may be utilized as a tool to detect early PI signs, such as ischemia. One rat study demonstrated that LSCI could capture the decrease in skin blood flow immediately after the release of mechanical load-induced tissue ischemia and before the occurrence of tissue necrosis.⁴¹ However, this device has not been studied in PI in humans at this time.

A colorimeter is a noninvasive imaging device that measures the color at the region of interest. A light source, such as xenon arc lamp, is used to illuminate the region of interest and a photodetector collects reflected light in the red, green, and blue regions of the visible light spectrum (∼600 nm, 560 nm, and 450 nm, respectively). The color is then calculated using the International Commission on Illumination’s L*a*b* (CIELAB) color space for objective expression of colors.⁴² This device has been used to detect erythema and melanin on the skin.⁴³ By quantifying the difference in skin color within the region of interest, early PI signs can be identified by skin color changes.^44–46

Similarities and Differences Across Devices

As with the clinical devices described previously, there is a common theme running through the 3 research devices—measurements of underlying variables associated with PI formation or occurrence, that is, ischemia and erythema. These research-based devices are objective, noninvasive and provide quantitative measures for the users in digital formats. Both LDF and LSCI quantify blood flow and can be used to visualize tissue ischemia at the skin level. The difference between the two is the measurement mechanism (frequency shift vs speckle contrast) and the size of the region of interest (single point vs whole limb). The measurement depth for LDF range between 0.5 mm and 1 mm,⁴⁷ whereas majority of the detected LSCI are from the top 700 μm of tissue.⁴⁸ All devices are FDA exempt; both LDF and LSCI are class II in the cardiovascular area, whereas colorimeter is class I and falls in the clinical chemistry and clinical toxicology devices.

These research-based devices provide measurement to visualize the “sign” of physiological change before PI formation. These measurements provide objective information that has the potential to be utilized by the clinicians to assist with decision making similar to that of the clinical devices reviewed in this paper; however, more research and development will be needed for interpretation of the measurements, and field test will be needed to determine its effectiveness for clinical use. The devices are not yet validated or approved for clinical implementation. Both LDF and LSCI share the limitation that is consistent across all blood flow monitoring technologies. Specifically, measurements are in arbitrary units, and comparisons across devices, or even across individuals without a reference, can be challenging. However, they have the potential to be used in a similar manner to clinical devices (eg, LWIT, SEM) in which measurements in the region of concern are compared with a control region. In contrast, colorimeter measurements of erythema and melanin are based on CIELAB color measurements,⁴² and there is ample evidence of research on CIELAB color space for erythema and eumelanin.^49–52 It should be noted, however, that not all commercial devices use the same scale for their melanin index and their erythema index and a conversion factor will be required for direct comparisons.

DISCUSSION

End User Priorities for PI Early Detection

Early detection of tissue damage can enable proactive measures to limit further development of an open wound, which is costly to treat. In a setting where health care resources are limited, effective use of a clinical device for PI early detection may allow for more targeted PI prevention strategies based on the objective data obtained from these devices. This ultimately can increase the likelihood of achieving beneficial outcomes for individuals with early signs of PI development.

Mapping the Devices and the PI Etiology

The current etiology of PIs includes direct deformation damage, inflammation, and ischemia.²¹ The devices included in this review has the potential to provide nonvisible information on tissue viability corresponding to the damage cascade. Localized inflammatory edema, that follows deformation-induced cell death may be monitored through SEM assessments.⁵³ Ultrasound imaging can also be used to detect hypoechoic areas indicating edema in soft tissue.⁵⁴ Relative temperature differences through LWIT may inform inflammatory or ischemic conditions. The NIRS, LDF, LSCI, and LWIT have the potential to further visualize ischemic conditions and deterioration of skin and underlying tissue before visible and tactile manifestation of PI damage. For the tissue death phase, ultrasonography could be used to capture the tissue structure changes corresponding to this phase, whereas a colorimeter could be used for measuring erythema for this phase. Because it would not be feasible in the clinical environment to have all clinical devices available on the market to target each and every one of these phases, the clinician will need to make an informed decision on the selection of device that will provide the best clinical data for their need (eg, patient population). Depending on the physiological attribute that the clinician wants to measure, there are devices available on the market for that purpose.

Implication for Use

None of the devices reviewed is meant to replace the standard skin assessment involving visual inspection and palpation for PI detection. For any patient admitted to the hospital, PI risk assessment should be conducted to identify the patients at increased PI risk. To continue caring for those patients at high risk, early detection devices can then be implemented in addition to the standard of care skin assessment to provide objective measurement of structural or physiological changes underneath the skin during routine skin inspection. When certain physiological changes are suspected by the clinicians, such as inflammation, ischemia, edema, these devices can provide valuable information to assist with further intervention. Early intervention can mitigate damage from the progression of a PI. Employing these technologies on admission to a facility, can help identify nonvisible signs and symptoms of any stage of PIs, including but not limited to a DTPI, and document as present on admission. In addition, focused interventions can be tailored to the needs of the individual.

Potential to Overcome Skin Tones’ Impact on PI Early Detection

Disparities and biases in health care are present due to ethnicity and race, as well as to skin tone itself.⁵⁵ Studies have shown that ethnicity or race is not an objective determinant of skin tone.^56–58 Individuals with dark skin tones are more likely to have more severe PI stages as compared with those with light skin tones.⁵⁹ One possible explanation of such a phenomenon was the missed opportunities in identifying Stage 1 PI in individuals with dark skin tones.⁵⁹ In the past 3 decades, the definition of Stage 1 PI has evolved and highlighted the differences in the non-blanchable erythema in individuals with dark pigmented skin with more details such as changes in sensation, temperature or firmness.⁶ Early detection devices included in this review have the potential to address this concern.

When using LWIT to measure skin temperature, Charlton et al reported that skin pigmentation has limited impact on the heat emitted (emissivity) from the skin.⁶⁰ Therefore, skin tone is unlikely to be an issue when assessing relative temperature differentials (ie, comparing adjacent tissue temperature to the target tissue temperature), which removes the influence of any factors that impact an individual’s absolute temperature (ie, a single temperature on a scale). The use of LWIT to identify relative temperature differentials in a region holds promise for overcoming missed identification opportunities associated in patients with dark skin tones.

The SEM demonstrated early detection in individuals with dark skin tones. A study comparing visual assessment of early PI and SEM measurement in nursing home residents showed that the incidence of PI 1 week later is associated with an elevated SEM value in both individuals with light and dark skin tones based on ethnicity.⁶¹ In the surgical intensive care populations, clinicians indicated that SEM assessments enabled more accurate skin tissue assessments in patients with dark skin tones.⁶² In a critical care unit, a 100% reduction in HAPI incidence was reported in the implementation period (N = 50), including 35 patients with varying dark skin tones.⁶³

Development of Phantom (Test Object) for Standard Evaluation of Technology

Developing performance evaluation procedures with standardized parameters would allow for the comparison between different early detection technologies. An ideal construct for such a standard would include PI models that could be used to test a device’s ability to detect simulated damage. Validated models would be able to mimic different sizes of PIs, at different stages, depths, and, or tissue types. The authors are unaware of any type of phantom that might adequately simulate all of the diverse parameters (temperature, moisture, blood flow, edema, tissue and blood oxygenation) simultaneously. An alternative strategy would be to develop test methods to assess the specific parameters within a category of early detection devices, so that products with the same category could be compared. This alternative strategy may be achievable and could build on existing test methodologies for individual technologies. For example, test methods for medical thermography measurement (eg, IEC 60601-2-59) and ultrasound systems (eg, IEC 60601-2-37) are well developed. Additional test methods could focus on evaluating characteristics directly related to clinical application, such as ease of use in clinical settings, integration with existing workflows, or interpretability of data by health care providers.

Dual Measurements and Machine Learning for Postprocessing of Measurements

An exciting clinical opportunity and future technology path also exists that would allow for combinations of different measurements using different techniques and devices to be combined within a “wound registry” that increases the effectiveness of these individual diagnostic procedures. For example, patients measured with multiple devices could be tracked. This allows the benefits of one detection technology to help develop the effectiveness of another technology by combining outcome determinations. This approach can help to further the effectiveness of PI detection technology, identify trends across patient types including those in minorities or with specific medical conditions and hence help provide the opportunity for improved PI outcomes for all.

A future area of technology and product development lies in leveraging digital health information (eg, EHR) and artificial intelligence/machine learning (AI/ML) models for bedside assessments of skin and underlying tissue. Emerging evidence demonstrated promising results of utilizing AI/ML based on readily available EHR data to predict HAPI development in surgical care patients⁶⁴ and the risk of HAPI in hospitalized patients.⁶⁵ Findings from these 2 research studies demonstrate the clinical utility and the potential for deploying ML-based algorithms in PI management workflows. Given that multiple technologies and devices now leverage cloud-based applications and digital EHR integrations, ML-based application that predict impending PIs shows promise. For example, Lustig et al⁶⁶ developed a novel ML algorithm for the prediction of heel DTPIs through SEM assessments reporting a strong predictive power in forecasting heel deep tissue injury events the next day, with sensitivity and specificity of 77% and 80%, respectively. Similarly, Asare-Baiden et al⁶⁷ used thermal and optical images using a forward looking infrared devices to explore the potential of ML-based thermography models to detect temperature changes in individuals with darker skin tones.⁶⁷ Further research and development will allow a wide range of nonvisible biomarker-based algorithms that may assist bedside clinical staff in achieving both earlier and accurate detection of developing PIs.

While the application of AI/ML continues to grow across health care, integrating these technologies into PI prevention workflows presents distinct opportunities and challenges. Critical challenges include ensuring the interpretability and transparency of ML outputs, mitigating the risks of automation bias and overreliance on algorithmic recommendations, and addressing the operational impact of false positives and negatives. Cost-effectiveness will need to be further explored given the financial considerations described extensively for the clinical devices. In addition, the ethical use of AI/ML in bedside decision-making and the development of reimbursement and governance structures remain key areas for future work. Thus, the focus should be on building responsible, scalable AI/ML frameworks that enhance the utility of available technologies.

CONCLUSIONS

The 4 clinical devices included in this review (ultrasonography, LWIT, SEM, and NIRS) measure different biological changes as a sign of PI before tissue breakdown, including tissue structural changes, inflammation/ischemia, edema, and tissue oxygenation, respectively. Although CPT codes are available for ultrasonography and NIRS, there is currently no CPT code for PI early detection for these devices. The 3 research devices included in this review (LDF, LSCI, and colorimeter) measure blood flow at a single spot or microvascular function, blood flow of an entire limb and differences in color of a region of interest, respectively. There is currently no standard to compare the performance of the above-mentioned devices on PI early detection, however, standards for each category achievable in the future. Current evidence supports the use of these devices and technologies as adjunct strategies to support standard skin assessments for patients at high risk of PI development, especially on admission to a facility. Further research on the reliability and validity of nonvisible evidence on early PI detection and the effectiveness of interventions to prevent or mitigate tissue destruction is warranted.