Skip to main content

A Retrospective, Descriptive Analysis of Hospital-acquired Deep Tissue Injuries

Empirical Studies

A Retrospective, Descriptive Analysis of Hospital-acquired Deep Tissue Injuries

Index: Ostomy Wound Manage. 2018;64(11):30-41.


Preventing, identifying, and treating deep tissue injury (DTI) remains a challenge. Purpose: The purpose of the current research was to describe the characteristics of DTIs and patient/care variables that may affect their development and outcomes at the time of hospital discharge. Methods: A retrospective, descriptive, single-site cohort study of electronic medical records was conducted between October 1, 2010, and September 30, 2012, to identify common demographic, intrinsic (eg, mobility status, medical comorbidities, and incontinence), extrinsic (ie, surgical and procedural events, medical devices, head-of-bed elevation), and care and treatment factors related to outcomes of hospital-acquired DTIs; additional data points related to DTI development or descriptive of the sample (Braden Scale scores and subscale scores, hospital length of stay [LOS], intensive care unit [ICU] LOS, days from admission to DTI, time in the operating room, serum albumin levels, support surfaces/specialty beds, and DTI locations) also were retrieved. DTI healing outcomes, grouped by resolved, partial-thickness/stable, and full-thickness/unstageable, and 30 main patient/treatment variables were analyzed using Kruskal-Wallis, chi-squared, and Fischer exact tests. Results: One hundred, seventy-nine (179) DTIs occurred in 141 adult patients (132 in men, 47 in women; mean patient age 64 [range 19–94]). Of those patients, 110 had a history of peripheral vascular disease and 122 had hypertension. Sixty-nine (69) DTIs were documented in patients who died within 1 year of occurrence. Most common DTI sites were the coccyx (47 [26%]) and heel (42 [23%]); 41 (22%) were device-related. Median hospital LOS was 23 (range 4–258) days and median ICU LOS was 12 (range 1–173) days; 40 DTIs were identified before surgery and 120 after a diagnostic or therapeutic procedure. Data for DTI outcome groups at hospital discharge included 28 resolved, 131 partial-thickness/stable, and 20 full-thickness/unstageable; factors significantly different between outcome groups included mechanical ventilation (15/42/12; P = .01), use of a feeding tube (15/46/12; P = .02), anemia (14/30/9; P = .005), history of cerebrovascular accident (12/27/7; P = .03), hospital LOS (67/18/37.5; P <.001), ICU LOS (23/10/12; P = .03), time-to-event (13.5/8/9; P = .001), vasopressor use after DTI (13/31/11; P = .003), low-air-loss surface (10/9/3; P = .005), and device-related (14/24/4; P = .002). Conclusion: DTI risk factors mirrored those of other PUs, but progression to full-thickness injury was not inevitable. Early and frequent assessment and timely intervention may help prevent DTI progression.


Despite progress in pressure ulcer (PU) knowledge during the past decade, the identification and treatment of deep tissue injury (DTI) continues to present a challenge for bedside clinicians. At the time this study was conducted, the definition of a DTI from the National Pressure Ulcer Advisory Panel (NPUAP) was “Purple or maroon localized area of discolored intact skin or blood filled blister due to damage of underlying soft tissue from pressure and/or shear. The area may be preceded by tissue that is painful, firm, mushy, boggy, warmer, or cooler as compared to adjacent tissue. DTI may be difficult to detect in individuals with dark skin tones. Evolution may include a thin blister over a dark wound bed. The wound may further evolve and become covered by thin eschar. Evolution may be rapid exposing additional layers of tissue even with optimal treatment.”1

In their descriptive article, Black et al2 described conditions not considered to be DTIs. Although the conditions the authors reported included intact skin that appeared purple or maroon, they also developed a process for differential diagnosis of the condition. Of note: During the period covered by the current study, the term pressure ulcer was the official term utilized in clinical practice and cited research likewise utilizes that term. As such, the terms pressure ulcer and deep tissue injury are used in this study. In 2016, the NPUAP revised the terminology from pressure ulcer to pressure injury (PI) to encompass the spectrum of tissue damage due to pressure. Since 2017, the Mayo Clinic (the employer of all authors) has adopted the updated NPUAP nomenclature.1,2

The accurate and timely identification of DTI is important for several reasons. Based on anecdotal experience, the current authors have found that early discovery of DTI allows prompt identification of possible causes, initiation of treatment, and potential development of preventive strategies. In addition, 24 to 72 hours can lapse between the precipitating pressure event and the onset of purple or maroon skin.3 This delayed manifestation becomes particularly important when the precipitating event occurred before the patient’s admission, yet the DTI appears beyond the 24-hour window for present-on-admission status, at which point the admitting facility becomes financially responsible for care. In Minnesota, for example, the law requires that this complication be publicly reported.4


In 2005 and 2013, the NPUAP consensus conferences focused on DTI in order to gain a deeper understanding into the causes, manifestations, and evolution of these injuries. A substantial advance was the understanding of DTI as more of a bottom-up than top-down phenomenon that was related to the deformation of deep tissue with muscle ischemia and reperfusion injury under intact skin, despite adequate surface pressure relief.5-9 The effectiveness (or lack thereof) of nursing interventions on the development of DTIs and their outcomes is highly relevant to wound, ostomy, and continence (WOC) nursing because the WOC nurse is usually the person consulted to assess and collaboratively manage these wounds over time.

Several recent studies describe characteristics of patients with DTI.9-11 In a large, annual, inpatient prevalence study, VanGilder et al10 surveyed between 79 000 and 92 000 patients from 2006 to 2009. The survey showed the overall and nosocomial PU prevalence decreased by approximately 1% in 2009 after remaining fairly constant in the years 2006 to 2008. However, the proportion of suspected DTIs increased 3-fold to 9% of all observed PUs in 2009; these wounds were more prevalent than either Stage 3 or Stage 4 PUs, with the heels the most prevalent DTI site. The authors of the study10 acknowledged that differences in demographic characteristics (that were not available to them) might help clarify the cause of more suspected DTIs in the study and provide insights to guide prevention, treatment, and design of future clinical studies. A prospective, multisite, exploratory study by Richbourg et al9 described the progression from suspected DTI to full-thickness skin loss and explored associated conditions of a sample of 40 inpatients. In this smaller study,9 the sacrum was shown to be the most frequent site of DTIs.

Since 2001, the NPUAP and others have made a substantial effort toward clarifying and refining the definition, clinical characteristics, and causes of DTI.12 As part of a retrospective review of a 2-year cohort of patients who developed DTIs, Sullivan13 reviewed and described patient factors contributing to the evolution and outcomes of these injuries.

In current study authors’ institution, DTIs are the second most common documented type of PU behind Stage 2 PUs. Although DTIs can be devastating injuries, progression to a full-thickness PU is not inevitable; DTIs have been observed to heal or improve before hospital discharge. The purpose of the current research was to describe characteristics associated with the development of DTIs and to better understand how these characteristics might be associated with PU outcomes at the time of hospital discharge. The authors elected to evaluate each DTI rather than each patient as the variable of interest, because patients with multiple DTIs did not acquire them all at the same time or in the same manner, nor did the injuries all resolve to the same degree.

Research questions were: 1) What are common demographic and intrinsic factors (eg, mobility status, medical comorbidities) of patients who develop a DTI during hospitalization? 2) What are common extrinsic factors (ie, surgical and procedural events, medical devices, head-of-bed [HOB] elevation, incontinence) of patients who develop a DTI during hospitalization? 3) What is the care and treatment applied to DTIs that resolve, improve by hospital discharge, or progress to a full-thickness PU?


This single-site, descriptive, retrospective analysis of electronic health records (EHRs) was approved by the Mayo Clinic Institutional Review Board (Rochester, NY). The study setting was a 2207-bed Midwestern academic medical center composed of a 2-campus acute care hospital designated by the American College of Surgeons as a level 1 trauma center. Inclusion criteria were documentation of DTI in hospitalized, adult patients who were 18 years or older at the time of admission, hospitalized for >1 day, and who had provided research authorization. Exclusion criteria were existing DTIs at admission, DTI in patients with a length of stay (LOS) <1 hospital day, and patients who had not agreed to review of their EHRs for research purposes. Suspected DTIs were excluded from the study analysis after the chart was reviewed by the study team, including a board-certified rehabilitation physician/physiatrist.

Additional data points related to DTI development or descriptive of the sample (age, gender, body mass index [BMI], race, comorbidities, Braden Scale scores and subscale scores, hospital LOS, intensive care unit [ICU] LOS, days from admission to DTI, time in the operating room [OR], serum albumin levels, support surfaces/specialty beds, and DTI locations) also were retrieved from the EHR on each patient by Mayo’s Division of Biomedical Statistics and Informatics. Information on other factors thought to be related to PI or DTI development (ie, devices, incontinence, vasopressor use, procedures exclusive of the OR, OR duration, comorbidities) and factors used in treatment following DTI occurrence (ie, dressings, noncontact low-frequency ultrasonic [NLFU] therapy) were manually retrieved by the study team for analysis. Specialty information related to physical therapy or the OR was examined by the appropriate specialty investigator. Several factors (ie, mechanical ventilation, need for a feeding tube) were used as surrogates for prolonged HOB elevation (≥30˚).14 The authors did not attempt to (nor could they) accurately capture degrees of HOB elevation from a retrospective chart review; however, from clinical experience, these factors could be used as rough surrogates for elevated HOB.

Data analysis. All data were extracted from the EHR and entered into an Excel (Microsoft Corp, Redmond, WA) database for statistical analysis. Statistical analyses were based on individual DTIs as the variable of interest because each was considered as an independent injury. For patients with multiple DTIs, not all DTIs occurred under the same circumstances or on the same date or resolved to the same degree.

Continuous features (age, hospital and ICU LOS, time to event, and Braden Scale for Pressure Risk Score) were summarized with medians, interquartile ranges (IQRs), and ranges. Categorical features (gender, race/ethnicity, and comorbidities) were summarized with frequency counts and percentages.

The outcome categories at time of hospital discharge were resolved, stable, partial-thickness, full-thickness, and unstageable. Because the small numbers of patients in 5 subgroups made meaningful statistical comparisons impossible, the subgroups were combined into 3 larger categories: resolved(defined as no evidence of DTI by EHR documentation), stable/partial-thickness (DTI fading toward normal pigmentation, no margin expansion at time of discharge by photo or EHR documentation, or epidermal loss not extending through  the dermis), and full-thickness/unstageable (progression to Stage 3, Stage 4, or unstageable requiring reporting to Minnesota Adverse Health Care Events by photo or EHR documentation).

Comparisons of factors between DTIs that resolved, remained stable/became partial-thickness, or progressed to full-thickness/unstageable by hospital discharge were evaluated using Kruskal-Wallis, chi-squared, and Fisher exact tests. All analyses were performed using the SAS software package, version 9.3 (SAS Institute Inc, Cary, NC). P values <.05 were considered statistically significant.


Between October 1, 2010, and September 30, 2012, 928 patients were documented with DTIs; of these, 151 were excluded because patients had not provided research authorization. An additional 592 DTIs were excluded as being present on admission or not verified by photo or note by an Advanced Practice Registered Nurse Clinical Nurse Specialist (APRN CNS), along with 6 additional ulcers lacking documentation. After the exclusions, 179 documented DTIs remained and were used as the cohort for analysis in this study.

Patient data. Of the 141-patient cohort (median age 64 [range 19–94] years), 25 (18%) had multiple DTIs: 16 had 2 injuries, 7 had 3 injuries, and 2 had 5 injuries. Characteristics of the cohort of patients who developed DTIs are described in Table 1, and the summary of patient characteristics (demographics and intrinsic and extrinsic factors) relative to DTI outcomes is shown in Table 2A, Table 2B, and Table 2C. The median BMI was 27.0 (range 12.7–61.6), and the median total Braden scores were 14 (range 8–23) at 3 days, 13 (range 8–22) at 2 days, and 13 (range 7–23) at 1 day before DTI appearance. The median hospital LOS was 23 (range 4–258) days and median ICU LOS was 12 (range 1–173) days, with a median time from admission to DTI event of 9 (range 1–129) days. The primary comorbid condition was hypertension (122, 68%), followed by peripheral vascular disease (110, 61%), congestive heart failure (71, 40%), diabetes (56, 31%), and acute renal failure requiring dialysis (56, 31%). Other factors that were identified and are known to contribute to DTI development and progression included altered mobility measured by a Braden subscale score of <2 (127, 76%), presence of incontinence (109, 66%), surgical intervention before DTI (40, 22%), diagnostic and/or therapeutic procedures requiring limited mobility (120, 66%), and use of medical devices (41, 23%).

Braden total and subscale scores were assessed 72, 48, and 24 hours before DTIs were identified. At all 3 times, the median total scores showed a high risk for PU development, with only a minimal downward trend. At 72 hours before DTI, the median total score was 14 (range 8–23), and at 48 and 24 hours the median total score was 13 (range 48 hours, 8–22; 24 hours, 7–23). No statistically significant difference was noted in the Braden total or subscale scores 72, 48, or 24 hours before DTI was identified, with the exception of the nutrition subscale score.

DTI data. Among the 179 DTIs studied, 47 (26%) occurred on women and 132 (74%) on men; 162 (91%) occurred on non-Hispanic whites, 7 (4%) on Hispanic whites, 2 (1%) on American Indian/Alaskan Natives, 2 (1%) on blacks, 1 (<1%) on an Asian, and 2 (1%) “other.” Two (2) patients who had a total of 3 (1.5%) DTIs did not disclose their race.

DTIs occurred in a variety of locations, including the sacrum (11, 6%), intergluteal (4, 2%), buttock (25, 14%), trochanter (1, 1%), ischium (1, 1%), calcaneal (42, 23%), coccyx (47, 26%), and other (48, 27%). There were no mucous membrane injuries in the “other” category.

Sixty-eight (68) DTIs occurred in patients who had received some care in the ICU, with a median ICU stay of 12 (range 1–173) days. Sixty-nine (69) of the DTIs occurred in patients who died within 1 year of the injury, with 38 of these DTIs occurring in patients who died during hospitalization.

Among the 179 DTIs, 41 (22%) were device-related: 16 (39%) were related to compression wraps, 1 (2%) to a sequential compression device, 2 (5%) to a walking splint-boot, 5 (12%) to a cast or splint, 3 (7%) to a cervical collar, 5 (12%) each to oxygen devices and orthotic braces, 3 (7%), to a bed/chair, and 1 (2%) to a sling.

NLFU MIST therapy (Cellularity, Inc, Warren NJ) was used for 56 (31%) DTIs, 2 before and 54 after DTIs developed. The therapy was used before a DTI developed for patients with a partial-thickness wound that subsequently developed additional tissue damage with DTI occurring postoperatively or post-procedurally.

Thirteen (13) DTIs were on patients who had some form of documented skin alteration at the same anatomic location before the DTI was identified (mean 77.5 hours; median 49 [range 20–192] hours). Of the 13 skin alterations documented, 8 were erythema, 3 were Stage 1, and 2 were Stage 2 PUs.

For 118 (66%) of the DTIs, patients had undergone some type of diagnostic or therapeutic procedure within 72 hours of DTI development, including computed tomography scan (38, 32%), radiography (35, 30%), ultrasonography (27, 23%), echocardiography (22, 19%), and interventional radiologic procedures (15, 13%).

DTI outcomes. At the time of hospital discharge, 28 DTIs had resolved, 131 were partial-thickness/stable, and 20 were full-thickness/unstageable. There were no statistically significant differences between the demographic characteristics related to the DTI outcome groups. Factors that did significantly differ between outcome groups included the intrinsic factors of a history of cerebrovascular accident (12/27/7; P = .03) and anemia after DTI (14/30/9; P = .005). Extrinsic factors included mechanical ventilation before DTI (15/42/12; P = .01) and after DTI 15/34/12 (P < .01), use of a feeding tube after DTI (15/46/12; P = .02), hospital LOS (67/18/37.5; P <.001), ICU LOS (23/10/13; P = .03), time to event (13.5/8/9; P = .001), vasopressor use after DTI (13/31/11; P = .003), low-air-loss surface after DTI (10/9/3; P = .005), and device-related injury (14/24/4; P = .002). (It is important to recall mechanical ventilation and feeding tube use were surrogates for HOB elevation to 30° or more.) All other intrinsic and extrinsic factors were not significantly different between the outcome groups, including surgeries and procedures.

In the patients who died, the DTIs appeared between 0 to 63 days before death. Although some of these may have been Kennedy Terminal Ulcers, insufficient documentation of the characteristics (eg, ulcer shape, borders, color) precluded the ability to confirm this diagnosis.15 A comparison of variables relative to DTI outcomes is summarized in Table 2A, Table 2B, and Table 2C.


This study was begun just before publication of the retrospective review by Sullivan13 of suspected DTI evolution in adult acute care patients, and the results share several similarities. Similar to Sullivan, the initial intent of this study was to describe and understand the characteristics of patients who developed a DTI and the contributing intrinsic and extrinsic factors involved. The long-term intent was to use this knowledge to inform clinical practice and future research. The current retrospective study examined factors associated with the development and outcomes of DTI during a 2-year period. Although no definite intrinsic or extrinsic factors could be directly attributed to the development of DTI, incidence of DTIs was higher in patients with cardiovascular diseases (including hypertension, peripheral and cerebrovascular disease, congestive heart failure) and diabetes mellitus.

In the current cohort of DTIs, (11%) progressed to full-thickness/unstageable at the time of discharge. The majority (131, 73%) were stable or were categorized as sloughing of the epidermis at the time of patient discharge. Twenty-eight (28, 16%) of the DTIs were completely resolved; 69 DTIs (38%) occurred in patients who died at a median of 51 days after the DTI was first identified, 38 (55%) of whom were still hospitalized. Cohorts in this and previously published work13 both had similar patterns of comorbidities, with a high incidence of patients who had cardiac, vascular, and/or renal diseases or diabetes mellitus.

Pre-DTI identification assessment. In 13 (7%) of the 179 DTIs, a skin alteration was documented before the DTI developed, and most of these (8 DTIs) initially showed blanchable erythema. The range of time that erythema was present varied greatly (ie, from 20 hours to 192 hours). For 3 DTIs, the PU was initially identified as Stage 1 and was present for 22 to 29 hours before being documented as a DTI. Two (2) DTIs were initially identified as Stage 2, with 1 present 25 hours before the DTI was documented. In this case, the patient underwent a procedure in which he was positioned over the Stage 2 PU; 48 hours after the procedure, the wound bed became dark purple and nonblanchable based on physical exam by the APRN CNS. Farid et al11 suggested blanchable erythema develops 168 to 336 hours before DTI formation, a range much higher than that of the current study. In the authors’ EHR system, the CNS on the unit was automatically alerted of skin alterations documented in the EHR to trigger the need for further assessment and intervention, which may explain this difference.

Chart information of the patients with the 13 DTIs and pre-existing skin alterations was reviewed to determine the types of diagnostic and surgical procedures associated with injury development. In a pilot study by Honaker et al,16 precipitating events occurred 1 to 5 (mean 2.41) days before DTI skin changes were noted. Therefore, the current study reviewed records for the 5 days before DTI documentation and identified 7 DTIs in patients who had surgical procedures lasting from 30 minutes to 5 hours. For 1 DTI, erythema was present for 168 hours before the DTI was identified, and the patient had undergone 7 diagnostic procedures in the 5 days before the DTI was documented. Another had erythema documented for 49 hours before DTI appearance; this patient had undergone 5 diagnostic procedures within that period of time.

Surgical and procedural considerations. In a surgical environment, DTI prevention involves identifying patients at risk through a skin assessment and reliably implementing prevention strategies. In the inpatient setting in the United States, the Braden Scale is a widely used tool for identifying at-risk surgical patients.17 However, in an OR environment, the preoperative Braden Scale score is inherently skewed because most patients have lost the ability for self-protection and communication. The Scott Triggers tool was developed as a predictive scale developed on evidence-based factors specifically tailored for high-risk perioperative PI or DTI development.17 The Munro Tool for Pressure Ulcer Risk-Assessment Scale for Perioperative Patients was developed as both a communication tool and documentation of risk assessment in each phase of perioperative care.18 The Munro Tool incorporates 15 evidence-based risk factors for the perioperative period. However, reliability and validity testing have not been completed on either of these tools. Good hand-off communication and visual indicators can alert OR staff as to which patients are most at risk because of preoperative instability and diagnostic procedures, as well as specialty knowledge of required positioning and equipment used during the case. If a patient develops a DTI postoperatively, root cause analysis needs to include OR or procedural environment staff, so all involved personnel can examine and possibly improve their practice.19 Although the differences between the outcome groups in the current study were not statistically significant, more postoperative DTIs evolved to full-thickness/unstageable PIs postoperatively (7 DTIs, 35%), and there was a higher incidence of procedures before DTI appearance.

Treatments and NFLU therapy. No discernible pattern of dressings used on DTIs, either before or after the DTI, was identified, although silicone-border dressings were used most frequently (37, 95%); these dressings were used before DTI occurrence in 20 DTIs. Use of specialty beds increased after DTIs appeared, with the greatest increased use in the group with resolved outcomes (P = .002). The only specialty bed used that showed significance between DTI outcome groups was a powered, multizoned, low-air-loss mattress system (10/9/3; P  = .005) compared with (partial) high-air-loss mattress (1/8/3; P  = 0.45).

DTI has been shown in a review of the literature by Berlowitz and Brienza20 to result from tissue distortion related to vertical and shear forces; in addition, the review by Stekelenburg et al21 and the rat model study by Cui et al22 demonstrated that compression of the blood supply, impaired lymphatic function of the tissues, and reperfusion injury also can be factors. Several investigators including Peirce et al (rat model)23 and Oomens et al24,25 and Agam and Gefen26 (biomedical engineering models) showed damage from oxygen-derived free radicals released during reperfusion can cause inflammatory and cytotoxic effects on the tissues, rendering muscle tissue more susceptible to this damage than the skin. NLFU provided at 40 kHz has been shown by multiple investigators27-31 to improve healing by decreasing proinflammatory cytokines, increasing vascular endothelial growth factors and improving microcirculation. In retrospective studies by Honaker et al31 and Thomas,32 an average of 10 to 12 sessions were required to produce a 75% to 100% resolution rate of DTIs. During the time frame covered by this study, NLFU was used for patients who did not have measurable signs of healing with standard treatments including dressings, specialty beds, and turn/reposition programs. The mean number of sessions for the NLFU group in the current series was 9.9, but 33% of the patients had 5 or fewer sessions because the DTI resolved or the patient was discharged or died. To achieve the best outcomes with NLFU, the intervention should be started as soon as the DTI is identified. In the current study, the average delay between DTI identification and NLFU initiation was 2 days, but in 9 DTIs (16%), there were delays ≥5 days.

Unlike the results of the prospective randomized controlled trial by Prather et al27 and a quasi-experimental study by McCormack and Hobbs33 that showed improved healing of partial-thickness injuries, the outcomes for the current cohort of DTIs that received NLFU were not better than those who did not. Several confounding factors were more prevalent in the NLFU-treated group, such as longer LOS (47 days vs 38 days), vasopressor use following DTI identification (38% vs 23%), and mechanical ventilation (50% vs 25%).

Although no specific patient acuity data were collected, the NLFU-treated group of DTIs were in patients who had a higher acuity of illness, because death occurred in 28% of the NLFU group within 1 month of discharge and in 70% of those patients, death occurred on the day of discharge or 2 to 3 days later.

HOB elevation. Since the introduction of guidelines for the prevention of ventilator-associated pneumonia, clinicians have had to weigh competing patient care priorities with regard to HOB elevation. Guidelines for ventilator-associated pneumonia34,35 recommend a HOB elevation between 30˚ and 45˚, and NPUAP guidelines recommend a HOB elevation of <30˚.36,37
A feasibility study by Schallom et al35 found reduced oral secretion volumes and reflux at higher HOB elevations without pressure-related tissue injuries. However, the 11 patients who completed the Schallom study35 all were on low-air-loss mattresses, and the study period was only 48 hours. In the current study, ventilator support was found to be a significant factor for PU development. This conflict in perceived priorities related to HOB elevation could be addressed with a more frequent turn-reposition program and use of specialty mattresses, such as those with low-air-loss, pressure-redistributing, or shear-reduction features.

LOS. In the current study, hospital LOS ranged from 4 to 258 days, with a median of 23 days, and ICU LOS ranged from 1–173 days with a median of 12 days. DTIs that were resolved at discharge had the longest ICU and hospital LOS, as well as the latest onset of DTI following admission. Most patients who had an extended LOS were awaiting transplant or dependent on supportive technology, such as extracorporeal membrane oxygenation or left ventricular assist devices. The authors’ analysis of DTI outcomes was limited to the status at hospital discharge, so although 131 DTIs (73%) were considered stable (not worsening) or improving (skin intact, discoloration fading), the ultimate DTI outcome was unknown. A total of 12 DTIs that resolved before discharge were device-related; the extended LOS due to more severe illness most likely allowed time for complete resolution.

Devices. In a presentation to the 2012 NPUAP Bi-Annual Conference, Baharestani38 noted that PUs caused by medical devices have increased in recent years. Medical device-related PUs are defined by the NPUAP1 as resulting “from the use of devices designed and applied for diagnostic and therapeutic purposes. The resultant pressure ulcer generally conforms to the pattern or shape of the device.” A secondary analysis of medical center data39 has shown medical devices create pressure against the skin, and the microclimate is altered as a result of humidity and heat between the device and the skin. This combination makes PUs more likely to occur after device use. To prevent these injuries, health care facilities have put policies in place to ensure medical devices are removed or repositioned in a timely manner; when a device is used, the NPUAP recommends that it be cushioned and the skin under a medical device inspected at least daily unless medically contraindicated and more often in patients with localized or generalized edema, and that staff be educated on use of medical devices.39

Of the 179 DTIs in the current study, 41 (23%) were related to medical devices. The device most likely to cause a DTI was a long-stretch compression wrap (Deluxe LF elastic bandage; Hartmann Inc, Rock Hill, SC) applied to the lower extremities. Wraps accounted for 15 (38%) of the DTIs caused by medical devices. As a result of this finding, the authors changed practice to use of short-stretch wraps (Rosidal K; Lohmann & Rauscher, Milwaukee, WI) for ambulatory patients and long-stretch wraps for nonambulatory patients. Cotton or foam padding was applied before wrap use to protect bony prominences and tendons from focused pressure. Casts and splints accounted for 8 (18%) of the device-related DTIs.

Other medical devices that caused DTIs in this study included cervical spine collars (3, 7%), cast/splint (5, 12%), respiratory devices (5, 12%), braces (5, 13%), boots (2, 5%), sequential compression devices (1, 2%), slings (1, 2%), and oscillating beds (3, 7%). In their prospective study, Coyer et al40 found ICU patients who developed a device-related DTI had an average of 8.6 medical devices placed. Endotracheal tubes and nasogastric tubes caused the highest number of DTIs in their study.

In the current study, 27 (68%) of the DTIs caused by medical devices occurred in the ICU setting and of those 11(49%) were caused by long-stretch wraps. A unique situation arises when patients are unconscious and require a cast or splint. If a patient cannot communicate that the cast or splint is causing additional pain, a PU may be more likely to occur. In rare cases the authors have experienced, removing the splint or cast routinely can place the patient at risk for further complications, such as worsening an unstable fracture.
Incontinence. According to a systematic review41 and a prospective analysis,42 fecal and urinary incontinence usually do not cause DTIs, but they have been associated with PUs because of the weakening of the skin’s ability to tolerate shear/friction, moisture, caustic drainage, and bacteria.37,38Gefen43 used mathematical and computational modelling to demonstrate the effects of wetness-related friction on shear loads, with resulting reduced strength of wet skin. In a systematic review and meta-analysis of incontinence-associated dermatitis and moisture as risk factors for PU development, Beeckman et al41 found significant associations between both urinary incontinence alone and double incontinence and PU development. In the current sample cohort, a low incidence of urinary incontinence was noted because patients were either continent or had an indwelling urinary catheter. However, patients with DTIs had an higher incidence of fecal incontinence in all outcome categories, ranging from 60% to 84%, with the highest percentage in the full-thickness/unstageable group.

Braden subscale scores. No statistical difference was found between the 3 condition-at-discharge groups in total Braden score 1, 2, and 3 days before DTI appearance. Each group’s total scores placed them in the high-risk category (12–14), with slightly lower scores in the group that developed full-thickness/unstageable DTI. The Braden nutrition subscale score was the only subscale that showed a significant difference between groups before DTI development, with the lowest scores found in the group that developed full-thickness PUs.

Although the prospective, quasi-experimental repeated measures study by Serpa and Santos44 did not find the Braden nutrition subscale score to be predictive for PU development in hospitalized patients, the current authors were not surprised to observe that DTIs progressed to full-thickness injuries and were associated with the lowest scores. A minimal downward trend was noted in other subscale scores from 72 to 24 hours before DTI, but the trend seemed to be consistent with an increasing acuity of illness and intensity of therapy and care, which may have been associated with decreased mobility and increased tissue distortion.


This study had several limitations. All data were abstracted retrospectively from the EHR, so granular details of DTI assessment and nursing documentation were limited. APRN CNS notes and photos were reviewed before DTI inclusion, but no interrater reliability was established. A natural history study without interventions was not conducted in that the 2-year period of data was derived from patients in a clinical practice where clinicians and providers were using various means to reduce the risk of PIs, which may have influenced the results.


This retrospective analysis of hospital data demonstrated that several factors are associated with development of DTI — namely, HOB elevation, use of therapeutic medical devices, and cerebrovascular diseases. What these factors may have in common is the potential for tissue distortion, which could contribute to additional unrecognized risk of tissue injury. When patients also experience hemodynamic instability requiring vasopressors or anemia, potential injury to tissue may be exacerbated.  

This study also demonstrated that DTIs do not inevitably progress to full-thickness reportable events, even in high-risk patients. A high index of suspicion when caring for patients with these factors should warrant more frequent monitoring of skin (especially under medical devices) and promote the earlier use of therapeutic support surfaces that redistribute pressure and reduce shear. Future DTI studies could involve concurrent or retrospective analysis of risk factors and associations of those factors in a descriptive or case control design. Inclusion of the role of microvascular disease and tissue distortion also should be considered for future research.


The authors thank and acknowledge Christine M. Lohse, MS, Department of Biomedical Statistics and Informatics, Mayo Clinic, Rochester, Minnesota, for performing the statistical analyses of the study data.


Dr. Tescher is the Clinical Nurse Specialist, Surgical/Trauma/CV Surgery ICU, the Hyperbaric Oxygen Therapy Unit and the Gonda Vascular Center; Ms. Thompson is the Clinical Nurse Specialist, Multi-Specialty ICU and Ambulatory Dermatology, Department of Nursing; Dr. McCormack is the Clinical Lead in Wound Care, Department of Physical Medicine and Rehabilitation; Dr. Bearden is a Nurse Manager, Surgical Services, Mayo Clinic Hospital Saint Marys Campus, Department of Nursing; Dr. Christopherson is a Consultant Physiatrist, Department of Physical Medicine and Rehabilitation; Ms. Mielke is the Clinical Nurse Specialist, Respiratory Care Unit and Medical ICU; and Ms. Sievers is the Clinical Nurse Specialist, Skin Population, Department of Nursing, Mayo Clinic, Rochester, Minnesota.


Please address correspondence to: Ann N. Tescher, APRN, CNS, PhD, Department of Nursing, Mayo Clinic, 200 First Street SW, Rochester, MN 55905; email: