A Multicenter, Comparative Study of Two Pressure-Redistribution Mattresses with Repositioning Intervals for Critical Care Patients : Advances in Skin & Wound Care

Secondary Logo

Journal Logo


A Multicenter, Comparative Study of Two Pressure-Redistribution Mattresses with Repositioning Intervals for Critical Care Patients

Jiang, Qixia MSN, RN, ET; Liu, Yuxiu MM; Yu, Hui MSN, RN; Song, Siping BSN, RN; Li, Guohong MPH, RN; Liu, Haiying BSN, RN; Zhou, Yi BSN, RN; Zhu, Yajun BSN, RN, ET; Jia, Jing BSN, RN, ET; Huang, Yunjuan MSN, RN; Wang, Jiandong PhD, MD

Author Information
Advances in Skin & Wound Care 33(3):p 1-9, March 2020. | DOI: 10.1097/01.ASW.0000653160.13611.5d
  • Free



In 2014, pressure injuries (PIs) were redefined as “localized injury to the skin and/or underlying tissue usually over a bony prominence as a result of pressure or pressure in combination with shear.”1 Critical care patients can easily develop PIs on bony areas of the body because of many risk factors such as sedation, mechanical ventilation, vasopressor support for BP and cardiac output, long periods of confined or restricted movement related to hemodynamic instability, and the inability to turn themselves.2 Further, some critical care patients have multiple organ dysfunction, which could cause cytokine release and initiate an inflammatory process, leading to edema and fluid overload in the skin: a PI waiting to happen.3 In addition, incontinence is common among critical care patients. If the skin remains in contact with the caustic substances in urine or feces for prolonged periods, it can become macerated, leading to the development of PIs.3

The incidence of PIs in trauma ICUs has been reported between 33% and 56%.4 Many recent studies have demonstrated variations in the prevalence and incidence of PIs in different countries and regions. In China and the US, the prevalence of PIs in critical care patients has been reported between 11.88% and 13%, and the incidence of PIs overall has been reported to be 4.48% to 10%.5–7 In Germany, the PI prevalence in 30 nursing homes and 13 hospitals ranged from 3.4% to 4.8%.8 The presence of any PI in critical care patients exacerbates their condition, requiring extended hospitalization and increasing the risk of mortality.9 Therefore, PI prevention must be a treatment priority.9,10

Research indicates that support surfaces and repositioning interval selection are very important for the prevention and management of PIs.11–14 Specifically, selection of the right surface for the right patient at the right time influences the effectiveness of PI prevention.15–17 However, there is little evidence to guide the selection of suitable pressure redistribution mattresses or repositioning intervals for at-risk patients.16,17 In light of this, the study authors conducted a multicenter study to evaluate the effectiveness of two pressure redistribution mattresses combined with two repositioning intervals for preventing PIs in a critical care patient population.



In Chinese hospitals, the types of pressure redistribution mattresses available include static foam mattresses and powered air-pressure redistribution mattresses. The powered air-pressure redistribution mattress used in this study (YQ-PBV Sanma; Sanhe Medical Equipment Co, Ltd, Shanghai, China) consists of nylon polyurethane with multiple chambers, an electric pump, an inflation valve, and a connecting tube. The electric pump continuously pushes air into multiple chambers of the mattress until the air pressure is sufficient to support the body. When someone lies down on the mattress, their body essentially floats on the air chambers, decreasing pressure on the skin and bony prominences.13

The viscoelastic (memory) foam mattress (Alpha-PLA-s; Taica Corporation International Trading Co, Ltd, Shanghai, China) used in this study consists of patented silica gel, two-layer moderate elastic foams, two-layer high elastic foams, and high-density polyurethane. The thickness of the mattress is approximately 12 cm. These components create a viscosity effect, redistributing the surface pressure under bony prominences and reducing the shear force and friction during patient repositioning.18–21

Setting and Participants

The study was carried out in 13 ICUs in 7 Chinese acute care hospitals with a total of 10,400 beds, including 243 ICU beds. Setting selection criteria included:

  1. an acute care hospital with more than 500 beds and a medical, surgical, and/or trauma ICU with at least 30 beds;
  2. at least one certified enterostomal therapist (ET) nurse on staff who was capable of managing the research in each hospital; and
  3. a hospital willing to participate and sign a joint research agreement.

After selecting and recruiting the participating hospitals, researchers obtained approval for the study from each hospital’s ethics committee and signed a joint research agreement.

In brief, participants were critical care patients who were hospitalized in the ICU and at risk of PI development.6 Inclusion criteria for participants included an age of at least 18 years and admission to a medical, surgical, or trauma ICU in the previous 24 hours. Other eligibility criteria included an expected length of stay of at least 7 days, as well as either limited activity and mobility (Braden Scale activity and mobility score of 1 or 2) or a risk of PI development (Braden Scale score ≤16).20,21 Patients had to be bedbound (activity score of 1) or wheelchair-bound (severely limited or nonexistent ability to walk, inability to bear one’s own weight, or requiring assistance getting into a chair or wheelchair; activity score of 2), and either be completely immobile (not making even slight changes in body or extremity position without assistance; mobility score of 1) or have very limited mobility (making occasional slight changes in body or extremity position but unable to make frequent or significant changes independently; mobility score of 2).21

Patients were excluded if they had any PIs at admission, including erythema or bruising; were in the terminal stage of an illness; were limited in or refused the use of a pressure redistribution mattress; required limited regular repositioning because of their medical condition; had mental disorders or psychiatric symptoms; or were participating in another study at the same time. Written informed consent for study participation was obtained from all enrolled patients or their family members.


The present study was a multicenter, open-label, comparative study conducted in seven acute care hospitals in China from July 2013 to July 2015. The trial group-to-control group allocation ratio was 1:1.

In clinical practice, patients are often repositioned every 2 or 4 hours when using suitable pressure redistribution mattresses.16,17 The protocol of 2-hour turns combined with a powered air-pressure redistribution mattress is popular in acute care hospitals in China; all participating hospitals used this protocol before the start of the study.13 Therefore, this protocol was used as the control condition in this study. For the trial group, the intervention involved a 4-hour repositioning interval combined with the use of a viscoelastic foam mattress, hereafter referred to as the test protocol. The reasons for selecting these protocols were as follows:

  1. The control protocol has been popular in most Chinese hospitals for over 30 years, and a previous multicenter study indicated that the protocol may be effective in preventing PIs.6
  2. Viscoelastic foam mattresses have been used in ICUs in some Chinese hospitals since 2012. A recent meta-analysis showed that an effective pressure-relief mattress combined with a 4-hour repositioning interval did not increase the incidence of PIs.22 In addition to submitting the two protocols to the hospital ethics committee for approval, a pre-experiment was carried out in seven hospitals for 1 month before the start of the study to ensure that they were reasonable and feasible. A total of 84 patients were randomly assigned to the two protocols (42 patients in each group), and no PIs occurred. After discussion by the research team, the two protocols were deemed safe and feasible for this population.
  3. According to the 2012 guideline developed by the Pan-Pacific Pressure Injury Alliance, repositioning frequency must consider the efficacy of the mattress. However, there is a great uncertainty regarding the repositioning intervals and selection of pressure redistribution mattresses.23 Therefore, the present study adopted a noninferiority design, with the aim of guiding clinical nursing care by comparing PI prevention between the two pressure redistribution protocols.

A research task force was formed, comprising nine certified ETs from the seven participating research hospitals. The task force developed a plan to train the ICU nurses in the participating hospitals before the research started. Topics covered in the training plan included selecting patients, applying the two pressure redistribution protocols and other preventive measures such as incontinence care and nutrition support, taking digital photographs of any problems that developed, sending the photographs to the ETs, caring for PIs with dressings, using the data collection/recording instrument, and assessing patients’ risk of developing PIs with the Braden Scale and incontinence scale.

To ensure the accuracy and reliability of the risk assessment with the Braden Scale, the research task force selected two nurses (assessing nurses) in each participating ICU and tested the accuracy of their Braden Scale risk assessments before the start of the research. The ETs in the seven participating research hospitals were responsible for testing the selected assessing nurses.

Sample Size

The primary objective of this study was to demonstrate the noninferiority of the test protocol compared with the control protocol in terms of PI incidence. Because large samples and reliable data on the incidence of PIs in China are lacking, the researchers referred to relevant international data when designing the primary outcome. For example, in 11 hospitals in six National Health Service trusts in England, Nixon and colleagues21 reported the incidence of stage 2 and worse PIs as 10.3% and 10.7%, respectively, when pressure redistribution support surfaces were used among patients 55 years or older who had been admitted to vascular, orthopedic, or medical wards.

Because stage 1 PIs were also included as an outcome in this study, the expected incidence was increased to 12%. The noninferiority margin was set at 4%, meaning that the test protocol would be clinically acceptable as noninferior to the control protocol if the PI incidence of the trial group was 16% or less. If each group included 817 cases, the statistical power needed to demonstrate such noninferiority would be 80% with a one-sided α level of .05. A 5% dropout/withdrawal rate for each group was assumed, yielding a required sample size of at least 860 cases per group.


Researchers used the stratified block randomization method with the research unit (ICU) as the stratification factor and a block size of four. Within each block, two patients were randomly assigned to the trial group, and the other two to the control group. A statistician not affiliated with the research generated the random allocation sequence.

Based on the results of the random allocation sequence, the head of the research task force prepared opaque envelopes with sequential numbers on the outside and a paper card inside that read “viscoelastic foam mattress + 4-hour repositioning protocol” or “powered air pressure redistribution mattress + 2-hour repositioning protocol.” These boxes were then delivered to the ICUs. The sealed envelopes were checked and managed by an ET in each hospital. Participating nurses opened the envelopes in sequence and assigned patients to the corresponding protocols. To reduce the risk of bias as much as possible, all patients knew that they would use a pressure redistribution mattress combined with a repositioning schedule, but they did not know which mattress or schedule they had. The two types of pressure redistribution mattresses were encased in the same type of cover and fitted with the same type of white cotton sheet (Nantong Xiaode Textile Co, Ltd, Nantong, Jiangsu, China).

To evaluate the main outcome measure without bias, a blind evaluation of skin problems among the patients was conducted during the study by certified ET nurses who did not know the patient’s grouping.


Patients received the control or test protocol until they were discharged from the hospital, died, or had remained in the ICU for at least 7 days. The participating nurses cooperated with medical doctors to create a repositioning schedule and a safety plan for the enrolled patients.

All enrolled patients were assessed for PI risk daily by qualified assessing nurses using the Braden Scale.3,21 In each participating ICU, two charge nurses who were not involved with the study were responsible for conducting examinations once per day of how the protocols were being carried out. These included checking patient positioning and compliance with the intervention, as well as the records kept by the participating ICU nurses. If these charge nurses found any problem, they would report it to the research task force, which would work to resolve the problem.

The participating nurses inspected the patients’ skin from head to toe at each repositioning during the intervention. If they found any skin problem, they took digital photographs and immediately sent the images to the ET in their hospital (again, all ETs were blinded to the patients’ group assignments). The ET then checked the photographs and skin problems at the bedside as soon as possible to confirm whether PIs were present and determine staging.24 If one ET could not confirm the presence of PIs, the photographs were sent to the research task force via a messaging app for timely discussion.

If a patient had a stage 1 or stage 2 PI, an ET would direct the participating nurses to manage it with a suitable polyurethane foam dressing to be changed twice per week.25 If a patient had a stage 3 PI or worse, an ET would treat the PI with the DIME (debridement, infection/inflammation, moisture balance, edge-nonhealing wound management) model for wound care and use negative-pressure wound therapy if necessary.25–27 If a patient with a PI was discharged from the ICU or hospital, an ET would follow up once a week in wound care clinic until 3 months to evaluate healing.

Outcomes Index

The primary outcome was the incidence of a new PIs during the intervention period. Secondary outcomes were the Braden Scale score, adverse events related to the protocols, time to development of a new PI, and extent of healing for existing PIs during the patient’s stay in the ICU and 3-month follow-up.

Data Collection Instrument

The research task force developed a case report form (CRF) based on previous multicenter research in China.13 The CRF included six categories:

  1. General information on the type of hospital and ICU. The number of beds in each hospital and ICU was collected at admission.
  2. Information on patient sex, age, medical diagnosis, skin trauma (yes or no), and expected length of stay in the ICU (collected on admission).
  3. Risk of PIs: assessed at admission and every morning using the Braden Scale. Because the Braden Scale does not specifically account for incontinence, the incontinence score from the Norton Scale was used to assess continence condition on admission.28 Data on serum albumin, hemoglobin, and blood glucose were also collected at admission to evaluate the risk of PIs indirectly.4
  4. A head-to-toe skin observation: conducted on admission and at every repositioning during the intervention.
  5. Specific PI observations such as location, development time, management methods, healed or unhealed, and healing time during the intervention period. Pressure injury stages were defined using the classification system developed by the National Pressure Injury Advisory Panel and European Pressure Ulcer Advisory Panel in 2009, including stage 1 to stage 4 and unstageable or suspected deep-tissue PIs (SDTPI).20
  6. Adverse events related to the mattresses or repositioning.

Completed CRFs were checked and signed by ETs and then delivered to the head of the research task force for data entry.

Statistical Analysis

EpiData software, version 3.0 (EpiData Association, Odense, Denmark), was used to establish the database, and SPSS for Windows, version 16.0 (SPSS Inc, Chicago, Illinois), was used for statistical analysis. The primary outcome analysis was based on the intention-to-treat principle. Cases with missing data on the primary outcome were omitted.

For the primary outcome, a two-sided 90% confidence interval (CI) was calculated for the difference between the two protocols (trial group minus control group). An upper limit of the 90% CI of less than 4% would indicate noninferiority. Other data were analyzed without imputation of missing values. Categorical variables were expressed as percentages and compared using Pearson χ2 test or Fisher exact test. Continuous variables were expressed as means, SDs, medians, and ranges. Comparisons between the two protocols were performed using the Student t test for normally distributed variables and the Mann-Whitney U test for non-normally distributed variables. Assessments of normality for the continuous variables were carried out using the Kolmogorov-Smirnov test. For continuous variables with repeated measurements (eg, Braden Scale scores), comparisons of different intervention times in each group were performed using the Friedman test. A two-sided P < .05 was considered statistically significant.

When 500 cases were completed in each group, a blinded interim analysis of the primary outcome was conducted without correcting for type I error. This analysis was not planned to infer noninferiority but to identify the incidence level of PIs and further adjust the research plan.


This research was approved by the medical ethics committee of Jinling Hospital, Nanjing University School of Medicine (no. 2013NYL014). Patient participation was voluntary. Each participant or family member was given information about the research protocol and aim, and they agreed to participate before enrollment. Investigators assured participants and their family members that their rights would be respected and guaranteed participant confidentiality. There was no penalty for study withdrawal at any time or refusal to participate in the research.


Researchers identified 1,654 eligible patients from July 2013 to December 2015. Of these, 1,204 patients were enrolled in the study after signed consent forms were obtained from 978 patients (81.2%) and 226 family members (18.8%). A total of 450 eligible patients were not enrolled in the study: 206 patients or their family members refused participation, 35 patients could not provide informed consent, and 209 patients had used other mattresses because of participation in other studies. All 180 participating nurses were compliant with the study protocols for mattress use, repositioning, and data collection. The completed CRFs were sent to the head of the research task force in a timely manner.

In accordance with the research plan, the study authors analyzed the interim data for 500 patients after each group completed observation and data collection. Of these patients, eight withdrew because the family members or patients decided to stop treatment (four patients from each group). Among 992 patients with complete data, 10 had a new single PI, an incidence of only 1%. This was far below the expected incidence. After discussion, the research task force decided that the research could not continue following the planned noninferiority design. The research task force therefore notified the involved hospitals to complete the observation for enrolled patients but to select no new cases, and the research ended.

At the end of the study, an additional three patients had new single PIs, and two additional patients (in the trial group) were excluded from the study because of missing data. Ultimately, 1,194 patients were included in the analysis, with 596 and 598 patients in the trial and control groups, respectively (Figure 1).

Figure 1.:

Of the 1,194 analyzed patients, 804 (67.3%) were in medical ICUs, having suffered a stroke, pulmonary infection, cardiac diseases, or renal diseases; 140 (11.7%) were in surgical ICUs after undergoing major surgery; and 250 (20.9%) were in trauma ICUs with soft tissue or bone trauma. There were 726 men (60.8%) and 468 women (39.2%) in the sample.

A total of 314 patients (26.3%) were ventilated, and 98 (8.2%) were sedated. Of the patients using ventilation, 160 (26.9%) were in the trial group, and 154 (27.8%) were in the control group. The difference was not statistically significant between the two groups (Fisher exact test; P = .358). Fifty-eight patients (9.7%) in the trial group and 40 (6.7%) in the control group were sedated. Similarly, this difference was not statistically significant between the two groups (Fisher exact test; P = .095).

In total, 918 patients (76.9%) were bedbound (Braden Scale activity score of 1), and 276 (23.1%) were wheelchair-bound (Braden Scale activity score of 2). In the trial and the control groups, 470 (78.9%) and 448 (74.9%) patients were bedbound, and 126 (21.1%) and 150 (25.1%) were wheelchair-bound, respectively. The difference was not statistically significant between the two groups (Fisher exact test; P = .114).

In the trial and the control groups respectively, 173 and 180 patients were completely immobile (Braden Scale mobility score of 1), 385 and 393 patients had very limited mobility (Braden Scale mobility score of 2), and 38 and 25 patients had slight limitations (Braden Scale mobility score of 3). The difference was not statistically significant between the two groups (Pearson χ2 test; P = .235). The patients’ characteristics at baseline are shown in the Table.


Pressure Injury Incidence, Characteristics, and Prognosis

Of the 1,194 patients in this study, 13 had single new PIs stage 2 or worse. These included nine men and four women, ranging in age from 39 to 88 years. The total PI incidence in the trial and control groups was 0.3% (2/596) and 1.8% (11/598), respectively. The difference between the two groups was 1.5% (95% CI, 0.2%-2.6%; P = .022).

The most common PI site was the sacrum, which accounted for 84.6% (11/13) of the observed PIs. The heel and occipital region accounted for 7.7% (1/13) of the PIs. The most common stage was stage 2, which accounted for 76.9% (10/13) of the PIs. Stage 4 and SDTPIs accounted for 15.4% and 7.7% (2/13 and 1/13) of the PIs, respectively. In the trial group, two patients had single stage 2 PIs, accounting for 100% (2/2) of the observed PIs in this group. In the control group, patients with stage 2 PIs accounted for 72.7% (8/11) of the observed PIs, and those with stage 4 PIs and SDTPIs accounted for 18.2% (2/11) and 9.1% (1/11), respectively. The Fisher exact test showed no significant differences between the two groups (P = .871).

The median time to PI development in the trial and control groups was 2 and 5 days (P = .231), respectively. The healing percentages of the PIs in the trial and control groups were 100% (2/2) and 45.5% (5/11), respectively. The median time of PIs healing in the trial and control groups was 6.5 and 7 days (P = .231), respectively.

Risk of Pressure Injury Formation

The 26 qualified assessing nurses in the 13 participating ICUs used the Braden Scale to complete daily assessments of the risk of PIs developing in the enrolled patients. The median Braden Scale score on admission was 13 for both the trial and control groups (range, 11–14 and 11–15, respectively), and the difference was not statistically significant between the two groups (Mann-Whitney U test; P = .151). During the intervention, there was a slow rise in Braden Scale scores from day 1 to day 7 for the two groups (median, 13; range, 11–15 and 11–15), which means that the risk of PIs decreased over time (Figure 2). Friedman tests showed a significant increase in the Braden scores at different intervention times within each group (P < .001), and the Mann-Whitney U tests showed no significant difference in these scores between the two groups at each time point.

Figure 2.:
BOX PLOT OF BRADEN SCALE SCORES BY GROUPThis box plots show the Braden Scale scores of each group at different interventional times. The lower and upper limits of the box indicate the first and third quartiles; the bar inside the box, the median; and the T bars indicate the 2.5 and 97.5 percentiles. Friedman tests show significant increase for the Braden scores at different times within each group (P < .001), and Mann-Whitney U tests show nonsignificant difference between two groups at each time.

Adverse Events

There were no reports of harmful or adverse events related to the two pressure redistribution protocols by the patients or nurses during the intervention period.


In this study, researchers selected ICU patients who were at risk of developing PIs (Braden Scale score ≤16) to use two pressure redistribution protocols to determine which protocol is better for preventing PIs. The results indicated that the total incidence of PIs was 1.1% (13/1,194). This was similar to a previous finding of 1% among 1,074 postoperative patients from 12 acute care hospitals using static and powered air pressure redistribution mattresses with repositioning every 2 hours for 5 days,13 and it was also close to Cullen’s10 report that incidence of PIs was less than 2% among hospitalized ICU patients. However, this finding was very different from the incidence of PIs between 10.3% and 10.7% found with the use of pressure redistribution mattresses and overlays in vascular, orthopedic, medical, and elderly care wards in six National Health Service trusts.21 It also was lower than the 4.48% incidence of PIs found among 1,094 ICU patients in a multicenter cross-sectional survey conducted in 12 Chinese acute care hospitals.6

Researchers hypothesize three possible reasons for the disagreement with prior studies. First, in the present study, all 180 participating nurses were compliant in receiving the two pressure redistribution protocols, and no complaints were recorded during the study period (Figure 1). This result was better than that of the multicenter cross-sectional survey, where, for patients at risk of developing PIs in 12 Chinese acute care hospitals, only 61.8% used the pressure redistribution mattresses and 75.1% were repositioned every 2 to 4 hours.13 This indicates that pressure redistribution mattresses and regular repositioning are very important for preventing PIs.

Second, investigators found that only 42.9% of the patients had a Braden Scale score of 13 or lower (Table). This percentage was smaller than what has previously been reported,5 which indicates that a Braden Scale cutoff score of 13 shows the best balance among sensitivity, specificity, positive predictive value, and negative predictive value for ICU patients. A lower total Braden score means a greater risk of developing PIs.20 Compared with critical care patients in other studies, Braden Scale scores observed in this study indicated that a lower percentage of patients were at risk of developing PIs and that fewer patients were on ventilation and under sedation. In other words, the conditions in Chinese ICUs differ from those previously reported in other countries.20,24

Third, no harmful or adverse events related to the two pressure redistribution mattresses or repositioning intervals were reported by patients or nurses during the study. This could indicate that the nurses were in good compliance with study protocol and the patients got good results.

Based on these results, the study authors believe that both pressure redistribution protocols are safe and effective for use among critical care patients. It was previously reported that viscoelastic foam mattress use among patients with existing PIs could improve healing and prevent new PIs by relieving pressure on bony prominences and skin surfaces effectively.18,19 Another study reported that the difference in PI incidence was not significant among three pressure redistribution protocols (high-density foam mattresses combined with 4-, 2-, or 3-hour repositioning) for older adult patients in 27 nursing homes, and it was especially notable that foam mattresses combined with a 4-hour repositioning protocol did not increase the incidence of PIs.12

Baseline data indicated that the trial and control groups were comparable (Table). At the end of the study, the incidence of PIs decreased 1.5 percentage points more in the trial group than in the control group (0.3% vs 1.8%, P < .05). This finding was similar to the result reported by Bergstrom et al.12 A possible reason for this finding was the effect of the material of the viscoelastic foam mattress. This type of mattress effectively disperses and decreases pressure on the bony prominences and skin surfaces, as has been previously reported.18,19 Researchers also found that using the viscoelastic foam mattress was comfortable for patients, and no adverse event reports related to allergies or allergic contact dermatitis occurred during this study.

There was insufficient information about the incidence of the primary outcome compared with the predicted data because researchers did not fully consider differences in the national context and patient characteristics. Consequently, the present study could not be completed using the planned noninferiority design. However, researchers did find that the incidence of PIs was lower in the trial group than the control group (P < .05), and there were no significant differences between the two groups in the time to development of PIs or the healing time of existing PIs (P > .05).

The boxplots revealed significant increases in the Braden scores at different intervention times within each group (P < .001), and the Mann-Whitney U tests showed no significant differences in these scores between the two groups at any time point (P > .05; Figure 2). This means that the 4-hour repositioning protocol combined with a viscoelastic foam mattress did not increase the risk of developing PIs or the incidence of PIs among ICU patients compared with the control protocol. This finding is similar to the results of Bergstrom et al.12,29 Using a 4-hour repositioning protocol combined with a viscoelastic foam mattress would be beneficial in terms of diminishing the nursing care burden, and study authors believe this protocol may be safe and viable to use in the prevention of PIs among critical care patients.


The intervention period was only 7 days because of the temporary nature of ICU stays. Twenty-eight patients in the trial group and 26 patients in the control group were discharged from the hospital after 7 days, and 408 patients in the trial group and 372 patients in the control group were transferred to another ward within 7 days because their conditions had improved. Therefore, long-term observations were limited, which could feasibly change the results.

In addition, the actual incidence of PIs was far lower than initially expected. Consequently, researchers had to stop the planned study midterm and could not complete the originally planned noninferiority evaluation. Because there are too few cases of PIs in this study (n = 13), stratified analysis and block comparisons were not possible.

Future studies should evaluate the effects of different repositioning intervals (every 2, 3, and 4 hours) and pressure redistribution mattresses on preventing PIs to optimize PI prevention protocols based on clinical needs.


Both tested pressure redistribution protocols can be used safely and effectively to prevent and manage PIs among critical care patients in Chinese acute care hospitals. The 4-hour repositioning protocol combined with a viscoelastic foam mattress was beneficial to ICU patients, did not increase the incidence or severity of PIs among these patients, and may be preferable because of the reduced nursing care burden. Ultimately, suitable pressure redistribution mattresses and repositioning intervals are very important to PI prevention. Evidence-based facility protocols enhance the awareness and capacity of nurses and patients to prevent PIs and increase the effectiveness of clinical nursing care.


1. Zhou Q, Yu T, Liu Y, et al. The prevalence and specific characteristics of hospitalised pressure ulcer patients: a multicentre cross-sectional study. J Clin Nurs 2018;27(3-4):694–704.
2. Edsberg LE, Black JM, Goldberg M, et al. Revised National Pressure Ulcer Advisory Panel pressure injury staging system: revised pressure injury staging system. J Wound Ostomy Continence Nurs 2016;43(6):585–97.
3. Estilo ME, Angeles A, Perez T, Hernandez M, Valdez M. Pressure ulcers in the intensive care unit: new perspectives on an old problem. Crit Care Nurse 2012;32(3):65–70.
4. Wilczweski P, Grimm D, Gianakis A, et al. Risk factors associated with pressure ulcer development in critically ill traumatic spinal cord injury patients. J Trauma Nurs 2012;19(1):5–10.
5. Cooper KL. Evidence-based prevention of pressure ulcers in the intensive care unit. Crit Care Nurse 2013;33(6):57–66.
6. Jiang Q, Li X, Qu X, et al. The incidence, risk factors and characteristics of pressure ulcers in hospitalized patients in China. Int J Clin Exp Pathol 2014;7(5):2587–94.
7. Lyder CH, Wang Y, Metersky M, et al. Hospital-acquired pressure ulcers: results from the national Medicare Patient Safety Monitoring System study. J Am Geriatr Soc 2012;60(9):1603–8.
8. Lechner A, Lahmann N, Neumann K, Blume-Peytavi U, Kottner J. Dry skin and pressure ulcer risk: a multi-center cross-sectional prevalence study in German hospitals and nursing homes. Int J Nurs Stud 2017;73:63–9.
9. McGinnis E, Greenwood DC, Nelson EA, Nixon J. A prospective cohort study of prognostic factors for the healing of heel pressure ulcers. Age Ageing 2014;43(2):267–71.
10. Cullen GE. Reducing hospital acquired pressure ulcers in intensive care. BMJ Qual Improv Rep 2015;4(1).
11. National Pressure Ulcer Advisory Panel. Support Surface Standards Initiative: Terms and Definitions Related to Support Surfaces. December 2018. https://cdn.ymaws.com/npiap.com/resource/resmgr/s3i_terms-and-defs-feb-5-201.pdf. Last accessed January 2, 2020.
12. Bergstrom N, Horn SD, Rapp MP, et al. Turning for Ulcer ReductioN: a multisite randomized clinical trial in nursing homes. J Am Geriatr Soc 2013;61(10):1705–13.
13. Jiang Q, Li X, Zhang A, et al. Multicenter comparison of the efficacy on prevention of pressure ulcer in postoperative patients between two types of pressure-relieving mattresses in China. Int J Clin Exp Med 2014;7(9):2820–7.
14. Norton L, Coutts P, Sibbald RG. Beds: practical pressure management for surfaces/mattresses. Adv Skin Wound Care 2011;24(7):324–32.
15. Black J, Berke C, Urzendowski G. Pressure ulcer incidence and progression in critically ill subjects: influence of low air loss mattress versus a powered air pressure redistribution mattress. J Wound Ostomy Continence Nurs 2012;39(3):267–73.
16. Chou R, Dana T, Bougatsos C, et al. Pressure ulcer risk assessment and prevention: a systematic comparative effectiveness review. Ann Intern Med 2013;159(1):28–38.
17. McNichol L, Watts C, Mackey D, Beitz JM, Gray M. Identifying the right surface for the right patient at the right time: generation and content validation of an algorithm for support surface selection. J Wound Ostomy Continence Nurs 2015;42(1):19–37.
18. Swafford K, Culpepper R, Dunn C. Use of a comprehensive program to reduce the incidence of hospital-acquired pressure ulcers in an intensive care unit. Am J Crit Care 2016;25(2):152–5.
19. Tomova-Simitchieva T, Lichterfeld-Kottner A, Blume-Peytavi U, Kottner J. Comparing the effects of 3 different pressure ulcer prevention support surfaces on the structure and function of heel and sacral skin: an exploratory cross-over trial. Int Wound J 2018;15(3):429–37.
20. Hyun S, Vermillion B, Newton C, et al. Predictive validity of the Braden Scale for patients in intensive care units. Am J Crit Care 2013;22(6):514–20.
21. Nixon J, Cranny G, Iglesias C, et al. Randomised, controlled trial of alternating pressure mattresses compared with alternating pressure overlays for the prevention of pressure ulcers: PRESSURE (Pressure Relieving Support Surfaces) trial. BMJ 2006;332(7555):1413.
22. Yuhong Z, Qixia J, Yanxia G, et al. A meta-analysis of repositioning frequency for preventing pressure ulcers among patients lying in pressure-relief mattress. Chinese J Nurs 2015;50(9):1029–36.
23. Gillespie BM, Chaboyer WP, Mclnnes E, et al. Repositioning for pressure ulcer prevention in adults [review]. Cochrane Database Syst Rev 2014:CD009958.
24. Jackson M, McKenney T, Drumm J, et al. Pressure ulcer prevention in high-risk postoperative cardiovascular patients. Crit Care Nurse 2011;31(4):44–53.
25. Guillen-Sola M, Soler MA, Tomas-Vidal AM. A multi-center, randomized, clinical trial comparing adhesive polyurethane foam dressing and adhesive hydrocolloid dressing in patients with grade II pressure ulcers in primary care and nursing homes. BMC Fam Pract 2013;14:196.
26. Ashby RL, Dumville JC, Soares MO, et al. A pilot randomised controlled trial of negative pressure wound therapy to treat grade III/IV pressure ulcers [ISRCTN69032034]. Trials 2012;13:119.
27. Baranoski S, Ayello EA. Wound dressings: an evolving art and science. Adv Skin Wound Care 2012;25(2):87–92; quiz 92-4.
28. Goldstone LA, Goldstone J. The Norton score: an early warning of pressure sores? J Adv Nurs 1982:7(5):419–26.
29. Demarre L, Verhaeghe S, Van Hecke A, et al. The effectiveness of three types of alternating pressure air mattresses in the prevention of pressure ulcers in Belgian hospitals. Res Nurs Health 2013;36(5):439–52.

acute care; comparative effectiveness research; ICU; PI; pressure injury; pressure redistribution mattress; repositioning

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.