RCT indicates randomized controlled trial.
Reddy M, Gill SS, Rochon PA. Preventing Pressure Ulcers: A Systematic Review. JAMA. 2006;296(8):974-984. doi:10.1001/jama.296.8.974
Author Affiliations: Divisions of Geriatric Medicine and Dermatology (Dr Reddy), Department of Medicine (Drs Reddy and Rochon), and Baycrest Kunin-Lunenfeld Applied Research Unit (Dr Rochon), University of Toronto, Toronto, Ontario; Women's College Hospital, Toronto, and Hebrew Rehabilitation Center, Boston, Mass (Dr Reddy); Institute for Clinical Evaluative Sciences, Toronto (Drs Gill and Rochon); and Division of Geriatric Medicine, Departments of Medicine and Community Health and Epidemiology, Queen's University, Kingston, Ontario (Dr Gill).
Context Pressure ulcers are common in a variety of patient settings and are associated with adverse health outcomes and high treatment costs.
Objective To systematically review the evidence examining interventions to prevent pressure ulcers.
Data Sources and Study Selection MEDLINE, EMBASE, and CINAHL (from inception through June 2006) and Cochrane databases (through issue 1, 2006) were searched to identify relevant randomized controlled trials (RCTs). UMI Proquest Digital Dissertations, ISI Web of Science, and Cambridge Scientific Abstracts were also searched. All searches used the terms pressure ulcer, pressure sore, decubitus, bedsore, prevention, prophylactic, reduction, randomized, and clinical trials. Bibliographies of identified articles were further reviewed.
Data Synthesis Fifty-nine RCTs were selected. Interventions assessed in these studies were grouped into 3 categories, ie, those addressing impairments in mobility, nutrition, or skin health. Methodological quality for the RCTs was variable and generally suboptimal. Effective strategies that addressed impaired mobility included the use of support surfaces, mattress overlays on operating tables, and specialized foam and specialized sheepskin overlays. While repositioning is a mainstay in most pressure ulcer prevention protocols, there is insufficient evidence to recommend specific turning regimens for patients with impaired mobility. In patients with nutritional impairments, dietary supplements may be beneficial. The incremental benefit of specific topical agents over simple moisturizers for patients with impaired skin health is unclear.
Conclusions Given current evidence, using support surfaces, repositioning the patient, optimizing nutritional status, and moisturizing sacral skin are appropriate strategies to prevent pressure ulcers. Although a number of RCTs have evaluated preventive strategies for pressure ulcers, many of them had important methodological limitations. There is a need for well-designed RCTs that follow standard criteria for reporting nonpharmacological interventions and that provide data on cost-effectiveness for these interventions.
Pressure ulcers represent a common but potentially preventable condition seen most often in high-risk populations such as elderly persons and those with physical impairments.1 The epidemiology of pressure ulcers varies considerably by clinical setting, with incidence rates ranging from 0.4% to 38% in acute care, 2.2% to 23.9% in long-term care (LTC), 0% to 17% in home care.2 In US acute care facilities alone, an estimated 2.5 million pressure ulcers are treated each year.3 The development of pressure ulcers can interfere with functional recovery, may be complicated by pain and infection, and can contribute to excesses in hospital length of stay.4 The presence of pressure ulcers is a marker of poor overall prognosis and may contribute to premature mortality in some patients.5,6
In addition to these adverse health outcomes, the financial impact of treating pressure ulcers is substantial. A Dutch study found that costs associated with care of pressure ulcers were the third highest after those for cancer and cardiovascular diseases.7 The price of managing a single full-thickness pressure ulcer is as much as $70 000, and US expenditures for treating pressure ulcers have been estimated at $11 billion per year.8,9 The development of pressure ulcers may also have important legal consequences: failure to prevent pressure ulcers in LTC settings has resulted in increasing litigation, with settlements favoring LTC residents in up to 87% of cases.10 These consequences highlight the value of preventing pressure ulcers.
Pressure ulcers can be prevented in many cases, and a targeted preventive approach may be less costly than one focused on treatment of established ulcers.2,11 A variety of preventive approaches have been proposed, and we undertook a systematic review to evaluate the evidence supporting these interventions.
We searched MEDLINE, EMBASE, and CINAHL from inception through June 2006, and the Cochrane Database through issue 1, 2006, to identify relevant randomized controlled trials (RCTs). We also searched UMI Proquest Digital Dissertations, ISI Web of Science, and Cambridge Scientific Abstracts. We used the following search terms: pressure ulcer, pressure sore, decubitus, bedsore, prevention, prophylactic, reduction, randomized, and clinical trials. One author (M.R.) further reviewed the bibliographies of identified articles. Criteria for selection of studies included RCTs that reported objective, clinically relevant outcome measures (such as incidence of pressure ulcers). There were no restrictions on language, publication date, or setting.
We grouped RCTs into 3 categories based on whether the intervention being evaluated addressed impairments in mobility, nutrition, or skin health. These impairments have been identified in previous research as important risk factors for development of pressure ulcers.12
We also classified studies by setting (acute care, LTC, rehabilitation facility), since the prevalence of pressure ulcers varies considerably between settings.5,13- 15
Assessing the effectiveness of nonpharmacological treatments, such as those used to prevent pressure ulcers, creates unique methodological challenges. For example, it may be impossible to blind participants and clinicians to the intervention. To address these challenges, Boutron et al16 recently developed a checklist to evaluate a report of a nonpharmacological trial (CLEAR NPT). We determined the methodological quality of the RCTs included in this systematic review using 6 elements from the CLEAR NPT: (1) adequate allocation sequence generation (ie, use of an appropriate method to generate the sequence of randomization); (2) concealed treatment allocation; (3) adequate participant blinding (where participant blinding was possible); (4) adequate outcome assessor blinding; (5) consistent follow-up schedule; and (6) intent-to-treat analysis. (Further details are provided at http://www.bichat.inserm.fr/equipes/Emi0357/docs/usersguidelines.pdf). Allocation concealment and double-blinding are strongly related to treatment effects.17- 19 More intensive follow-up in one arm of a trial (eg, more frequent turning) could contribute to reductions in the development of pressure ulcers, even if this was not part of the intervention being evaluated. Therefore, we examined trials to ensure that similar follow-up schedules were present in each group, unless a trial specifically set out to determine the frequency with which an intervention was applied.
The search strategy identified 763 citations, from which 59 relevant RCTs were selected. A QUOROM (Quality of Reporting of Meta-analyses) flow diagram (Figure) shows an overview of the study selection process. The 59 selected studies enrolled a total of 13 845 patients: 9397 (67.9%) in acute care, 2367 (17.1%) in LTC, 333 (2.4%) in rehabilitation, and 1748 (12.6%) in mixed settings. Sample size was unclear for 1 trial.20
Fifty-one RCTs evaluated interventions for impaired mobility and included 11 551 patients: 7984 (69.1%) in acute care, 1866 (16.2%) in LTC, 333 (2.9%) in rehabilitation, and 1368 (11.8%) in mixed settings (Table 1, Table 2, Table 3, and Table 4). The length of follow-up ranged from 1 to 224 days; in 5 trials the length of follow-up was unclear.
Support Surfaces. Quiz Ref IDSpecialized support surfaces (such as mattresses, beds, and cushions) reduce or relieve the pressure that the patient's body weight exerts on skin and subcutaneous tissues as it presses against the surface of a bed or chair. If a patient's mobility is compromised and this interface pressure is not relieved, the pressure can lead to impaired circulation and ulcer formation. Forty-eight of the 59 RCTs in our sample examined the role of support surfaces in preventing pressure
ulcers (Table 1, Table 2, and Table 3).20- 67
Quiz Ref IDPressure-reducing surfaces may be either static support surfaces (such as mattresses or mattress overlays that are applied to the top of a mattress and filled with air, water, gel, foam, or a combination of these) or dynamic support surfaces (which mechanically vary the pressure beneath the patient and thereby reduce the duration of the applied pressure). Dynamic support surfaces include alternating-pressure mattresses, low-air-loss beds, and air-fluidized mattresses. Alternating-pressure mattresses produce alternating high and low pressures between the patient and mattress, thus diminishing the period of high pressure. Low-air-loss mattresses consist of air sacs through which warmed air passes. Air-fluidized mattresses contain silicone-coated beads that liquefy when air is pumped through them. Dynamic support surfaces are generally more expensive than static surfaces, with air-fluidized mattresses being the most expensive type of dynamic support surface.
In a well-designed study of 446 patients undergoing elective major surgery, Nixon et al28 demonstrated that specialized foam mattress overlays on operating tables decreased the incidence of postoperative pressure ulcers. In other settings, specialized foam (eg, convoluted foam, cubed foam) and specialized sheepskin (denser and thicker than regular sheepskin) overlays were the only surfaces that were consistently superior to standard hospital mattresses in reducing incidence of pressure ulcers.22,23,32- 34,36,40
Four RCTs examined various types of seat cushions for the prevention of pressure ulcers.24,40,41,44 Three studies examined specialized foam cushions, with 1 study using a standard cushion for comparison and the other 2 studies using another type of specialized foam for comparison. The incidence of pressure ulcers was no different in the intervention groups. One study compared a specialized foam cushion with a combination specialized foam and gel cushion and found the latter to be significantly more effective.40
Fourteen RCTs directly compared dynamic and static support surfaces.46- 59 The best-designed trial of these was conducted by Vanderwee et al,46 who studied 447 patients and found no difference in pressure ulcer incidence between dynamic and static support surfaces. Only 3 trials found that dynamic support surfaces were better than static support surfaces,48,50,53 and 1 of these trials did not report statistical significance.50 One trial directly compared dynamic, static, and standard support surfaces and found no difference between the dynamic and static support surfaces but found that both were better than standard surfaces.58 In a well-designed RCT of 1972 acute care patients, Nixon et al60 found no difference in the incidence of pressure ulcers when dynamic support surface mattress overlays were used instead of dynamic support surface mattresses. The mattresses cost more than the overlays, but an economic evaluation conducted alongside the trial suggested that the mattresses may be more cost-effective and are more acceptable to patients than the overlays.71
Three RCTs compared beds that turn and rotate the patient with standard hospital beds or standard intensive care unit (ICU) beds.65- 67 Standard ICU beds were not clearly defined in these studies, but ICU beds are usually dynamic support surfaces. Rotating beds offered no advantage in reducing pressure ulcer incidence as compared with either standard hospital beds or ICU beds.65- 67
Repositioning. Quiz Ref IDPatient repositioning is a mainstay in most pressure ulcer prevention protocols, which often recommend turning every 2 hours. The aim of repositioning, like that of specialized support surfaces, is to reduce or eliminate interface pressure and thereby maintain microcirculation to regions of the body at risk for pressure ulcers. We were able to identify only 2 trials that specifically evaluated repositioning strategies.68,69 These trials included 884 participants (46 in acute care and 838 in LTC) (Table 4).
Defloor et al68 investigated the effect of different turning regimens in a 4-week RCT involving 11 LTC facilities. They found that turning patients every 4 hours combined with the use of specialized foam mattresses significantly reduced the incidence of pressure ulcers compared with turning every 2 hours on standard hospital mattresses. However, the methodology of this study was limited: there was no information to indicate that patients were randomly allocated with concealed allocation, there was inadequate blinding of participants and outcome assessors, and no intent-to-treat analysis was performed. Furthermore, this study did not simply compare different repositioning schedules but rather combined different repositioning schedules with different support surfaces in the 2 comparison groups. Therefore, it is difficult to advocate turning patients every 4 hours rather than the standard of every 2 hours based on this study alone.
One RCT investigated the efficacy of different patient positions.69 This small study of 46 elderly inpatients examined the difference between the 30° tilt position (the placement of pillows under one buttock and under each leg, so that the sacrum and heels are not in contact with the support surface) vs standard patient positioning (90° side-lying). No significant difference in outcomes was found between the 2 groups.
Exercise and Treatment of Incontinence. Investigators in 1 trial70 addressed the risk factors of immobility and incontinence (fecal and urinary) by examining skin health outcomes of a combined exercise and incontinence intervention (Table 4). Individuals with incontinence were recruited from 4 facilities. In the intervention group, research staff provided exercise and incontinence care for 2 hours per day for 32 weeks. The control group received usual care from LTC staff. This multifaceted intervention did not reduce pressure ulcer incidence relative to usual care.70
Quality of RCTs Targeting Impaired Mobility. The quality of the 51 RCTs that examined impaired mobility was generally suboptimal (Table 1, Table 2, Table 3, and Table 4). Of the 51 studies, 25 (49.0%) adequately described the generation of random allocation sequences, and only 14 (27.4%) gave information that indicated patients were randomly allocated with concealed allocation. There was inadequate blinding of patients, but this may be difficult when studying interventions involving support surfaces, repositioning, or exercise and treatment of incontinence. We therefore did not include ratings for the CLEAR NPT criterion of adequate participant blinding in Table 1 and Table 2. In some cases, however, it may be feasible to have a blinded observer perform outcome assessments, but this was described in only 10 (19.6%) of the 51 studies. The follow-up schedules in the study groups were consistent in 40 (78.4%) of the 51 studies. Intent-to-treat analyses were performed in 14 (27.4%) of the 51 studies. Only 3 (6.3%) of the 48 studies examining the role of support surfaces fulfilled all 5 of the applicable criteria we selected from the CLEAR NPT checklist.24,28,69 Of these 3 studies, 2 were small: 1 had a sample size of 32 patients,24 and another enrolled only 46 patients.69 Small sample size was a potential limitation of many studies; the mean number of participants was 226 (range, 11-1972). Participants in RCTs represented heterogeneous populations (including patients from general medical and oncology wards, as well as a variety of subspecialty surgical services including orthopedics, vascular surgery, and cardiothoracic surgery wards).
Interventions Targeting Impaired Nutrition. Five RCTs targeted impaired nutrition and included a total of 1475 patients: 974 (66.0%) in acute care and 501 (34.0%) in LTC (Table 5).72- 76 The length of follow-up ranged from 14 to 182 days. The intervention for all 5 RCTs consisted of mixed nutritional supplements.
The relationship between nutritional intake and prevention of pressure ulcers is often assumed but is based on limited evidence. The only RCT to find that nutritional supplementation was beneficial was conducted by Bourdel-Marchasson et al.73 This was also the largest and best designed of the intervention trials targeting impaired nutrition, suggesting that the smaller trials may have reported negative outcomes because they were underpowered. The trial by Bourdel-Marchasson et al studied 672 critically ill inpatients older than 65 years and compared standard diet alone to standard diet plus 2 oral nutritional supplements per day. Patients in the control group had a relative risk of pressure ulcer development of 1.57 (95% confidence interval, 1.30-2.38; P = .04), compared with those in the intervention group.73
Quality of RCTs Targeting Impaired Nutrition. Several important methodological limitations were identified for the 5 RCTs that examined the efficacy of nutritional supplementation. None of the 5 studies provided information to indicate that patients were randomly allocated with concealed allocation. Only 1 of the studies provided adequate blinding of participants and outcome assessors. Three of the studies demonstrated consistent follow-up between study groups. Only 1 of the studies performed intent-to-treat analysis. None fulfilled more than 3 of the 6 CLEAR NPT criteria (Table 5).
Interventions Targeting Impaired Skin Health. Three RCTs targeted impaired skin health and included a total of 819 patients: 439 (53.6%) in acute care and 380 (46.4%) in mixed settings (Table 6).77- 79 The length of follow-up ranged from 21 to 30 days.
Dry sacral skin is a known risk factor for the development of pressure ulcers.12 All 3 RCTs examined specific topical agents; none evaluated simply moisturizing skin as an intervention.
Torra i Bou et al77 compared the effects of a hyperoxygenated fatty acid preparation with those of a placebo treatment. Fatty acids have been thought to protect against friction and pressure and also to reduce hyperproliferative skin growth. Pressure ulcer incidence during the study was 7.32% in the intervention group vs 17.37% in the placebo group (P≤.006). van der Cammen et al78 hypothesized that topical nicotinate could enhance subcutaneous vascular supply but did not find any benefits of topical nicotinate when compared with a lotion containing hexachlorophene, squalene, and allantoin. Green et al79 proposed that hexachlorophene could act as a bactericidal agent and that allantoin might stimulate cell proliferation and tissue growth. They suggested that a lotion containing hexachlorophene, squalene, and allantoin was superior to a simple moisturizing lotion, but they did not provide any measure of statistical significance for this finding.
Quality of RCTs Targeting Impaired Skin Health. Methodology for the 3 RCTs that examined skin health was limited, though all were double-blinded. None of the 3 studies gave information to indicate that patients were randomly allocated with concealed allocation. One of the studies showed consistency of follow-up between study groups. None performed intent-to-treat analysis. None fulfilled more than 3 of the 6 CLEAR NPT criteria.
We identified 59 RCTs evaluating interventions to prevent pressure ulcers. Our review suggests that the methodology for pressure ulcer prevention trials is suboptimal overall, although more recent studies have shown improvements in methodological quality.24,28,60,71 In pressure ulcer prevention trials, it is sometimes not feasible to ensure that participants are blinded, and other aspects of these trials may be difficult to standardize. To address these issues, we used the CLEAR NPT quality-rating guidelines developed specifically for nonpharmacological interventions.16 Only 3 of the 58 RCTs in this review fulfilled all of the criteria we selected from the CLEAR NPT checklist.24,28,69
Of the 59 trials, 43 (72.3%) took place in acute care settings. This seems appropriate, given that the majority of pressure ulcers (60%) develop during acute care hospitalizations.9
The trials reviewed were generally short, but follow-up ranged from 1 to 224 days. Although pressure ulcers can develop within 2 to 6 hours, the incidence of pressure ulcers has been found to rise with increasing duration of stay in LTC, and continues to rise for at least 2 years.80 While days or weeks of follow-up may be adequate for patients with reversible risk factors (eg, relatively healthy patients in perioperative settings), patients with indefinite immobility (eg, paraplegia) may require longer follow-up.
Mattress overlays on operating tables may decrease the incidence of postoperative pressure ulcers.28 For hospital inpatients, however, the choice may be different; although dynamic support surface mattresses are initially more expensive than dynamic support surface mattress overlays, inpatients prefer the mattresses and they may be more cost-effective than overlays in the long run.60,71 Specialized foam and specialized sheepskin overlays reduce pressure ulcer incidence compared with standard hospital mattresses.22,23,32- 34,36,40 The choice between dynamic support surfaces and static support surfaces such as specialized foam or sheepskin is not clear, as only a few of the RCTs that compared these interventions showed any difference in outcomes.46- 59 Costs may be an important factor to consider when choosing between these strategies.
Quiz Ref IDOn the basis of 1 RCT, it appears that use of nutritional supplements may be of benefit in the prevention of pressure ulcers, though which specific nutrients offer the best protection remains unclear.73 It seems reasonable to recommend consultation with a dietician for patients at risk of developing pressure ulcers to ensure adequate general nutrition.
Quiz Ref IDDry sacral skin is known to be a risk factor for developing pressure ulcers. Moisturizing skin is inexpensive and unlikely to be of harm, so it would be a reasonable strategy to implement to prevent pressure ulcers. The incremental benefit and cost-effectiveness of specific topical agents over simple moisturizers is unclear.77- 79
There is a mismatch between the high prevalence and costs associated with pressure ulcers and the amount of good-quality research focused on their prevention. The majority of RCTs we reviewed focused on support surfaces, though these are often some of the most expensive interventions to implement. The cost of support surfaces varies considerably, from less than $100 for some types of foam to more than $30 000 for certain specialty beds.81 Given the labor-intensive nature of nursing care and the costs associated with various products, considerable work still must be performed to adequately determine the overall cost-effectiveness of interventions to prevent pressure ulcers and the appropriate targeting of these interventions to match these high costs with those individuals most likely to derive benefit. Recent studies have started to formally examine cost-effectiveness in this area.71,82
Future studies should also attempt to define the interventions required to prevent pressure ulcers specifically among high-risk populations. Risk factors for development of pressure ulcers include being bed- or chair-bound, being unable to reposition without assistance, difficulty with ambulation, history of stroke, fecal incontinence (which is strongly correlated with immobility), low body weight, lymphopenia, difficulty with independent feeding, impaired nutritional intake, nonblanchable erythema of intact skin (ie, a stage 1 pressure ulcer), and dry sacral skin.1,3,12,83- 87 Advanced age has not been identified as an independent risk factor for pressure ulcers. The negative results of some studies may reflect the fact that interventions were not appropriately targeted. For example, nutritional supplements may be of limited benefit in people who are not malnourished. Strategies to reduce pressure ulcers should be directed toward high-risk patients, and focused interventions should be targeted to patients with deficiencies in the specific domain being investigated.
In addition to examining these focused interventions, future studies of pressure ulcer prevention may benefit from viewing pressure ulcers from a geriatric medicine perspective. Geriatric syndromes such as falls and urinary incontinence tend to develop when compensatory mechanisms are compromised by the accumulated effect of impairments in multiple domains.88 In the case of pressure ulcers, coexistent impairments in mobility, nutritional status, and skin health often conspire together to produce ulcers. Thus, multifactorial interventions delivered by a multidisciplinary team may prove effective in preventing pressure ulcers, similar to interventions used to prevent other geriatric syndromes.89
Although it may not be possible for patients to be blinded to treatment when studying interventions involving support surfaces, repositioning, or exercise and treatment of incontinence, it is often feasible to have a blinded observer perform outcome assessments. This would be achievable particularly for studies taking place in the operating room (because the blinded observer could perform an assessment immediately after the patient has been transferred to another surface), for patients who are relatively easy to transfer (so they can be moved to another surface during assessments), and for some mattress overlays and seat cushions. Blinding of observers is particularly difficult in studies of dynamic support surfaces (since they are electric and may move or make noise) or when patients are critically ill and cannot safely be moved to another surface for assessments.
Several guidelines on the prevention of pressure ulcers have been developed.90- 92 Unfortunately, many physicians and nurses report feeling that they lack education regarding pressure ulcer management, suggesting that guidelines are not reaching their intended audience.93,94 More effective resources should be expended on knowledge translation of existing evidence. Guidelines alone may not work that well.95,96 In addition, further well-designed RCTs that follow standard criteria for reporting nonpharmacological interventions (such as the CLEAR NPT checklist)16 are needed. Head-to-head comparisons of the most promising interventions are also required to evaluate which ones are the most effective.
One limitation of our review is that incomplete reporting in the RCTs might have influenced our assessment. However, available evidence suggests that what is reported about key features of a study generally reveal what is actually performed.97,98
We assessed study quality using selected features of a previously developed checklist.16 We believe that the key components of this checklist were used. In addition, there are several ways to define study quality.18,99 Recent research has concentrated on 2 main issues: which components of the quality assessment are predictive of valid results and what checklist best assesses quality. Despite the many quality scales and checklists that have been created, the optimal approach is still unclear.99- 101 To avoid arbitrary quality scoring, we simply recorded whether the various components of the checklist were reported in the RCTs that we reviewed.
The methodological quality of RCTs evaluating interventions to prevent pressure ulcers is suboptimal but provides some valuable information on which to base recommendations for effective approaches to prevent this common condition. Specifically, the most promising interventions are using appropriate support surfaces (mattress overlays on operating tables, specialized foam overlays, and specialized sheepskin overlays), optimizing nutritional status, and moisturizing sacral skin. Repositioning is a mainstay of ulcer prevention, but it is not known whether certain strategies have advantages over others.
Further well-designed RCTs that follow standard criteria for reporting nonpharmacological interventions are needed. In particular, given the heterogeneity of the study populations involved in the RCTs we reviewed, further study is needed to confirm the generalizability of these interventions' effectiveness to different patient populations and settings. The prospective collection of data on cost-effectiveness in such RCTs would provide valuable information.
Corresponding Author: Madhuri Reddy, MD, MSc, Hebrew Rehabilitation Center, Department of Medicine, Boston, Mass 02131 (email@example.com).
Author Contributions: Dr Reddy had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design; drafting of the manuscript: Reddy, Rochon.
Acquisition of data; analysis and interpretation of data; statistical analysis: Reddy, Gill.
Critical revision of the manuscript for important intellectual content: Reddy, Rochon, Gill.
Administrative, technical, or material support: Gill.
Study supervision: Rochon.
Financial Disclosures: None reported.
Funding/Support: Dr Rochon was supported by a Canadian Institutes of Health Research Investigator Award and by ICE grant HOA-80075. Dr Gill was supported by an Ontario Ministry of Health and Long-Term Care Career Scientist Award.
Role of the Sponsor: The funding organizations had no role in the design or conduct of the study; the collection, analysis, or interpretation of the data; or the preparation, review, or approval of the manuscript.
Acknowledgment: We thank Carvill Lo, BA, Clinical Research Assistant, Women's College Hospital, Toronto, Ontario, for her administrative support. Ms Lo received no financial compensation for this support.