Use of Air-Fluidized or Fluid Immersion Redistribution Support Surfaces for the Treatment of Stage 4 Pressure Injuries: A Case Series

Pressure injuries (PIs) remain a serious health care concern. A report by the Agency for Health Research and Quality shows that hospital-acquired pressure injuries (HAPIs) increased by 6% from 21.7 HAPIs per 1000 discharges in 2014 to 23 HAPIs per 1000 discharges in 2017.¹ Prevalence of severe HAPIs (stage 3, stage 4, deep tissue pressure injury, and unstageable) remained at about 1% of the acute care population from 2011 to 2016.² It is the authors’ belief that the lack of progress in preventing these significant complications may be due to an increase in the aging population, survival of more medically complex patients, or a combination of multiple factors. It is, however, imperative that severe PIs are managed as efficiently and effectively as possible. 

Caring for severe PIs is costly to the health care system. According to the Agency for Health Research and Quality, PIs cost from $9.1 to $11.6 billion per year in the United States.³ About 59% of these costs are disproportionately attributable to a relatively small number of patients with stage 3 and 4 full-thickness wounds, which occupy clinician time and hospital resources.⁴ The cost of an individual patient’s care ranges from $20,900 to $151,700 per PI. Medicare estimated in 2007 that each PI added $43,180 in costs to a hospital stay.⁵ 

Because long-term acute care hospitals (LTACHs) specialize in treating patients who may have more than 1 serious condition, but who may improve with time and care and return home, and who require long stays (on average, more than 25 days),⁶ these hospitals have a higher overall prevalence of PIs when compared with rehabilitation hospitals. Overall prevalence in the LTACH environment ranged from a high of 32.9% in 2006 to a low of 28.8% in 2015.⁷ HAPI prevalence in LTACHs ranged from a high of 9.0% (2006) to a low of 5.6% (2015), which was generally still higher than in most other care settings.⁷ For the above clinical reasons, care decisions should be evidence-based. 

In the authors’ practice, the majority of patients who are admitted with PIs to 5 LTACHs located in the southeast United States have had skin organ failure due to life-threatening injuries and complex medical/surgical challenges, or have wounds that have been unresponsive to care from other treating facilities. Patient care goals in these LTACHs are to stop PIs from worsening by increasing in size or stage, begin the healing process, and minimize the risk of additional PI occurrence. Care plans are multifactorial and must include the use of a specialty support surface.

Currently, there are very limited data available to make decisions regarding support surface selection. A consensus panel published a surface selection algorithm, but due to the variable performance across different support surfaces, definitive recommendations were not given.⁸ The National Pressure Injury Advisory Panel (NPIAP, formally NPUAP) have put forth a tremendous effort and created the Support Surfaces Standards Initiative, whose efforts were designed to delineate surface performance.⁹ This team of clinicians, scientists, and industry supporters have recently published test methods for support surfaces, and there is ongoing testing in the laboratory setting.⁹  However, there remains a void of published evidence in the clinical community as to which surface heals a PI faster than another surface.

The present study sought to provide data regarding PI healing to enable informed decision-making regarding support surfaces. This study evaluated PI healing rates on the following 2 specialty support surfaces, which were indicated for persons with stage 3 and stage 4 pressure injuries: 1) a fluid immersion system (FIS) continuous low-pressure, low-air-loss support surface (Dolphin FIS; Joerns Healthcare, Charlotte, NC) and 2) an air-fluidized therapy (AFT) bed (Envella Air Fluidized Therapy bed; Hill-Rom Inc, Batesville, IN). The AFT bed provides fluid support that is generated by tiny silicone-coated ceramic beads that are blown upward toward a cover sheet to create a fluid medium. The goal is to maximize envelopment of the body and significantly reduce shear, friction, and pressure as well as mechanical stress applied to the skin and subcutaneous tissue when compared with a nonfluid support surface.¹⁰ AFT also provides microclimate management.¹⁰ The FIS surface also aims to provide immersion. This surface uses air bladders to achieve lower interface pressure and immersion. According to Mendoza et al,¹¹ the FIS surface does not provide microclimate management features. 

To provide objective clinical evidence, the present study aimed to compare the healing rates of stage 3 and stage 4 PIs in patients on an AFT bed compared with FIS continuous low-pressure, low-air-loss support surface in the LTACH environment.

This randomized controlled trial was conducted at 2 locations within a LTACH system located in the southeast United States beginning April 2018 and with a planned end date of April 2019. The study aimed to enroll 60 patients (30 patients per treatment arm) using block randomization. However, due to slow enrollment, only 5 patients participated in the study. Patients were randomized to use either the AFT or FIS surface. The protocol and informed consent were reviewed and approved by the Chesapeake Institutional Review Board (IRB) (Pro00024278). Patients (or their legal authorized representative) who met inclusion/exclusion criteria  and provided written or verbal informed consent were enrolled. 

Inclusion criteria included 1) curative or maintenance treatment goals (ie, not palliative), 2) adherence to the standard clinical care for the admitted condition (ie, the patient was not refusing care), 3) age between 18 and 85 years, 4) anticipated length of stay of at least 3 to 4 weeks (from date of baseline assessment) in the treating LTACH, 5) at least 1 stage 3 or 4 PI located on the weight-bearing aspect of the truncal/pelvic region, and 6) written or verbal informed consent from the patients or their legal authorized representative. Exclusion criteria included 1) unstable spinal cord injury; 2) weight < 70 lb or > 350 lb; 3) primary PI had previous flaps or grafts with significant associated scarring, which was clinically thought to impede wound contracture; 4) primary study PI was located over the trochanteric head of the femur; 5) current chemotherapy or chemotherapy within the past 6 weeks; and 6) patients who required immunomodulating drugs.

Variables. At baseline, primary study PIs were identified per study inclusion criteria and exclusion criteria. For each PI, wound data were collected at baseline and weekly thereafter using a 3-dimensional wound measurement device, (eKare Inc, Fairfax, VA); electronic data were captured directly into the study database (ClinPlus EDC, Anju Software, Phoenix, AZ). The wound measurement device was a validated system used to capture wound size data (area and volume) using an artificial intelligence system.^12,13 In addition to wound healing data, the number and severity of new PIs that formed on each study participant during the study period were recorded on case reports and entered into the study database. 

Patients were followed until discharge from the LTACH or until they required a different mattress for their admitted condition, reached a maximum of 12 weeks of study participation, had their care goals changed to palliative goals, required surgical closure of their targeted PI, required acute care admission > 24 hours, required bioengineered skin substitutes for their targeted PI, or experienced healing of their targeted PIs. Patients who completed at least 1 weekly follow-up assessment were included in the analysis.

Data analysis. Volume (mm³) and area (mm²) of primary study PIs were obtained at baseline and patient discharge. Total percent change in wound volume or wound area were calculated by dividing the difference in baseline and discharge values by baseline values and multiplying by 100. To determine the rate of healing per day, the overall change in area or volume was divided by the total number of days the patient was on the study surface. To determine the rate of healing per week, the daily rate was multiplied by 7. 

Study interventions. Standard care as per LTACH policy and procedures was provided for all patients in addition to support surface selection (FIS or AFT) and additional study-related assessments, which included specific laboratory assessments. The protocol of care included repositioning per an individualized schedule based on the patient’s need and condition, but not greater than every 2 hours. Wound debridement, topical dressing protocol, and negative pressure wound therapy (NPWT) were carried out per the clinical practice of the treating physician. 

Enrollment into the study was difficult. In October 2018 the admission criteria for LTACHs changed in the Long-Term Care Hospital Prospective Payment System. The new criteria required patients who qualified for LTACH admission to have a hospital stay of 3 or more days in an intensive care unit setting immediately prior to admission to the LTACH.¹⁴ This change significantly affected our admissions of patients with PIs. In an attempt to improve study enrollment, a study amendment was implemented to update the inclusion and exclusion criteria. The original criteria excluded patients from participation who had osteomyelitis or injuries in which bone exposure was > 2 cm² in area. These criteria were removed from the protocol via an IRB-approved protocol amendment, allowing an increase of the patient population that may qualify for the study. However, the effect was minimal due to patient weight and location of PI.

After 16 months of very slow enrollment, the study was discontinued after enrolling a total of 5 patients; of these 5 patients, 1 patient requested to discontinue the study because the randomized surface was uncomfortable and 4 patients completed at least 1 week of study participation. Two (2) of these 4 patients used the AFT support surface, and 2 patients used the FIS support surface. These 4 patients are presented in the current case series. Table 1 shows demographics, limited medical histories, and admission laboratory values for all cases. Table 2 presents changes in area and volume for all cases.

Case 1. A 62-year-old White man with a body mass index (BMI) of 26.3 kg/m² was admitted to the LTACH with enterocutaneous fistulas and a stage 4 sacral PI (volume = 58.6 cm³; area = 68.5 cm²) (Figure 1). The patient was cognitively aware, had impaired mobility, and required full assistance to mobilize while in bed and transfer to a chair. The patient was unable to ambulate and had an ostomy. On admission, Braden score was 15, hemoglobin (HGB) was 11.1 g/dL, hematocrit (HCT) was 33.6%, pre-albumin (ALB) was 22.3 mg/dL, and C-reactive protein (CRP) was 2.9 mg/dL. The sacral PI had been previously surgically debrided 1 month prior to admission to the LTACH, and the patient had been treated with NPWT. The patient reported between moderate and severe pain (5 of 10, with 10 representing worst possible pain) from his PI. He was placed on an FIS support surface, and NPWT was continued. 

At week 5 of the study, his health deteriorated and his care goals were changed to palliative care; therefore, he discontinued the study. Platelets were low (15,000 mL), and the patient was bleeding from a midline abdominal wound and from his sacral stage 4 PI. The PI had a 13% increase in volume and 7% decrease in area compared with admission measurements (Figure 1 and Table 2). The patient’s speech was hard to understand, and he had nonverbal signs of pain. 

Case 2. An 83-year-old African American woman with a BMI of 34.1 kg/m² was admitted to the LTACH for an infected stage 4 sacral PI (volume = 50.1 cm³; area = 26.3 cm²). This patient was bed-bound and could not shift weight significantly on her own. She was confined to bed (requiring full assistance to mobilize in bed), cognitively impaired, and had fecal incontinence and a urinary catheter. Her admitting Braden score was 13. Significant laboratory values included a low HGB (8.9 g/dL) and low HCT (27.7%). Her ALB was 2 g/dL and pre-ALB was 22.9 mg/dL. CRP was 3.01 mg/dL. The PI was treated with NPWT on admission, and the patient was randomized to the FIS support surface. 

On day 20 the patient was discharged from the LTACH and therefore discontinued the study. Laboratory values were relatively unchanged. The sacral PI had decreased in volume (38.5%) and area (43%) (Table 2 and Figure 2). 

Case 3. A 73-year-old White man with a BMI of 38.5 kg/m² was admitted to the LTACH for an infected stage 4 PI on his right buttock and a stage 4 PI on his coccyx (volume = 140.9 cm³; area = 37.0 cm²); the coccyx PI was the targeted PI for this study (Figure 3). He had severe protein calorie malnutrition (ALB of 1.6 g/dL and pre-ALB of 13.6 mg/dL). He was admitted to the study with a Braden score of 14, HGB of 8.3 g/dL, HCT of 26.4%, CRP of 14 mg/dL, and in moderate pain (7 of 10). He was cognitively aware and required partial assistance to mobilize in bed as well as to ambulate. He was randomized to the AFT support surface and treated with NPWT and nutritional supplementation. 

At day 13, the area of the coccygeal PI decreased 23.8% and volume decreased 38.5% (Table 2 and Figure 3). The patient reported no pain from the PI and had a Braden score of 17. NPWT was discontinued. ALB was slightly better at 1.9 g/dL, pre-ALB was lower at 12.9 mg/dL, HGB higher at 10.0 g/dL, and HCT higher at 30.7%. CRP increased from 14.0 mg/dL at admission to 18.4 mg/dL at discharge.

Case 4. A 71-year-old White man with a BMI of 23.1 kg/m² was admitted to the LTACH for a stage 4 PI (volume = 41.1 cm³; area = 30.9 cm²). He required partial assistance to mobilize in bed and transfer. This patient was unable to ambulate and was cognitively aware. He was continent of stool and had a urinary catheter. Admitting Braden score was 15. Laboratory values were significant for HGB 10.1 g/dL, HCT 31%, ALB 2.1 g/dL, pre-ALB 17.5 mg/dL, and CRP 2.9 mg/dL. He was randomized to the AFT surface, and the PI was treated with NPWT.

On day 13, the patient was discharged from the LTACH. Laboratory values had improved, and the sacral PI had reduced 56% both in area and volume (Table 2 and Figure 4).

Enrollment was a challenge in the current study. The original study design planned for 60 patients; however, a total of 5 patients enrolled and only 4 patients completed study requirements before the study was discontinued. This experience underscores the challenges associated with conducting comparison studies to evaluate wound healing rates of chronic wounds such as PIs. To capture meaningful differences in healing data, weekly wound assessments must be performed on an adequate number of patients for an extended time to provide an opportunity for wounds to heal and/or resolve. Patients admitted to an LTACH have an average length of stay of 25 days,¹⁵ making it a plausible environment to conduct these types of studies. However, staff time is limited and patients may be discharged to other treating facilities before their wounds resolve or significantly heal. 

In addition to the above challenges, the inclusion and exclusion criteria of the current study were too restrictive. Broader enrollment criteria and cooperation between different care environments within the same hospital network could provide more study resources and a longer data collection period to provide this much needed comparison data. From a study perspective, this could involve multiple IRB submissions (as facilities are usually different business entities), multiple principle investigators, and multiple informed consenting processes. It is the authors’ belief that such studies can guide the decision-making process when hospitals are evaluating which support surface provides the most therapeutic value to patients with severe wounds. 

The current study included a fragile patient population. During the study, no additional PIs developed in patients using either the AFT or FIS surface. At the time of enrollment, all patients had stage 4 PIs and received NPWT over the course of the study. The 2 patients who were placed on the AFT surface met the care goals of their LTACH admission and clinically improved. One (1) of the patients on the FIS surface also met basic care goals but did not experience PI healing as fast as the AFT patients. The condition of the other FIS patient deteriorated, likely due to systemic complications. Although these observations are clinically important, the lack of sufficient study enrollment greatly limited the ability to draw significant conclusions from the available study data.

AFTs are characterized by a relatively high level of envelopment and immersion compared with other support surfaces.¹⁶ They also have a relatively high evaporative capacity and heat withdrawal, which addresses skin microclimate.16 Data from previous studies demonstrate that stage 3 and 4 PIs have faster healing rates by area on air-fluidized beds compared with powered-air beds.^17,18 

Other studies have compared the FIS surface to the AFT bed. Mendoza et al¹¹ reported interim results from 40 patients who participated in a randomized controlled study comparing postoperative flap management on the FIS surface compared to an AFT bed; early data demonstrated a flap dehiscence rate of 13% for major dehiscence and 40% for minor dehiscence with an FIS surface compared with 0% for both minor and major dehiscence on an AFT bed. Additionally, the occurence of maceration was 27% compared with 0% for AFT, and FIS patients had a longer length of stay (18.28 days vs 10.05 days for AFT). The authors indicated that the FIS surface did not provide microclimate management, which may have affected the complication rate. A very moist skin microclimate increases friction and shear, which may negatively impact healing, according to Shaked and Gefen¹⁹ as well as Schwartz et al.²⁰ 

It is well established that it is important to prevent and manage PIs using evidence-based recommendations.¹⁴ Patients who are receiving acute care and had a PI were reported as having higher mortality rates (9.1%) than patients who did not have PI (1.8%; odds ratio 5.08).²¹ They also were more likely to be discharged from acute care to post-acute settings (72.5% compared with 24.7%; odds ratio 3.116) rather than home.²² Additionally, severe PIs put the patient at risk of localized infection and systemic sepsis.²¹

Providing care for severe PIs requires multiple therapies and interdisciplinary care to help resolve the patient’s multifaceted problems so that the wound will ultimately close. However, as described in this case study, conducting the studies needed to elucidate the effects of care on PI healing is a formidable challenge that requires multicenter studies.

The present study did not achieve enrollment goals. The authors present the data that are available and recognize the limited generalizability of these data. 

The purpose of this study was to compare healing rates for patients with severe PIs on AFT and FIS support surfaces using a randomized control study design. Enrollment was challenging, and only 4 patients (2 AFT and 2 FIS) with stage 4 PIs completed the study. The PIs of all but 1 patient, whose overall health condition deteriorated, showed a reduction in volume and area ranging from 19.3 to 54.3 cm³ and 0.12 to 1.34 cm² per day, respectively. The results of this study are encouraging but mainly illustrate the difficulties of conducting studies to evaluate the effectiveness of PI treatment modalities. More research, using larger sample sizes, is needed to determine the therapeutic benefit of these measures to enable evidence-based decision-making.

The authors thank Catherine VanGilder (Hill-Rom, Inc) and Dr. LeAnn Phipps for providing references and guidance. 

Ms. Griffey is the wound care specialist for Noland Hospital Anniston, Anniston, AL. Ms. Barnes is a wound care consultant to Noland Health Services Hospital Division, 2 Advise, Inc, Birmingham, AL. Ms. Bardonner is chief clinical officer, Hospital Division, Noland Health Services, Birmingham AL. Address all correspondence to: Marta K. Bardonner, RN, BSN, CWOCN, Chief Clinical Officer, Hospital Division, 600 Corporate Parkway, Suite 100, Birmingham AL 35242; email: mbardonner@nolandhealth.com.