Treatment of Pressure Ulcers:  A Systematic Review FREE

Quiz Ref IDPressure ulcers are regions of localized damage to the skin and underlying tissues that usually develop over bony prominences such as the sacrum or heels.^1- 3 These lesions are an important source of suffering for patients and their caregivers. Pressure ulcer prevalence varies widely depending on patient factors (eg, age, physical impairments) and treatment setting.^4- 7

Treatment strategies for pressure ulcers can be both costly and complex. Hundreds of different mattresses and local wound care products are currently promoted,⁴ and few have been evaluated in randomized controlled trials (RCTs). It remains unclear which of the many available treatments promote the most effective healing of pressure ulcers.^8- 11

While several effective strategies to prevent pressure ulcers exist,⁶ many patients continue to develop them. This is especially true in high-risk settings such as acute care hospitals, in which patients have reduced mobility.^12- 13 Thus, clinicians require an understanding of effective treatment options. We examined the evidence supporting interventions for the treatment of pressure ulcers.

The databases of MEDLINE, EMBASE, and CINAHL were searched from inception through August 23, 2008, to identify relevant RCTs. The following search terms were used: pressure ulcer, pressure sore, decubitus, bedsore, chronic wound, treatment, therapy, management, randomized, and clinical trials. A hand search also was performed to identify any other articles. Inclusion criteria were RCTs published in the English language that reported objective, clinically relevant outcome measures such as healing rates or wound size. When the search was not limited to studies published in the English language, 2 non–English-language trials were found: one in Italian (294 participants)¹⁴ and another in Japanese (19 participants).¹⁵ Because few non–English-language trials were found and the total number of participants in these trials was small, this study was limited to RCTs published in the English language. Studies that evaluated chronic wounds other than pressure ulcers or assessed only adverse events or secondary outcomes (eg, pain) were excluded. There was too much clinical heterogeneity in the individual RCTs to permit meaningful pooling of the data in a meta-analysis.

Sources of funding were extracted from the trials using a method described by Als-Nielsen et al.¹⁶ We also determined if the RCTs reported any potential author conflicts of interest.

Information was extracted regarding participant age, population studied, and treatment setting. Trials used different terms to describe treatment settings. These terms were grouped as follows: acute care, long-term care, palliative care, rehabilitation, ambulatory care, and home care. Pressure ulcers at the beginning of each trial are described by stage unless the trial used different terminology (such as superficial or deep).

The included RCTs were categorized into 3 groups depending on whether they investigated the management of underlying contributing factors, the effects of local wound care, or adjunctive therapies. This approach was selected because wound specialists approach pressure ulcer management sequentially; first, reduce or eliminate underlying contributing factors (support surfaces and nutritional supplementation), then provide local wound care (wound dressings and biological agents), and finally consider adjunctive therapies (eg, vacuum therapy).¹⁷

A criterion standard for quantifying outcomes in ulcer healing has not been established.^18- 19 Surrogate end points (eg, amount of granulation tissue, degree of debridement, and bacterial burden) do not directly measure healing and may not correlate with healing.^20- 21 Measurement of wound surface area, including wound depth and undermining (ie, tunneling under the skin), is a reliable and valid method of assessing wound healing.¹⁹ Therefore, only studies that calculated wound size with wound volume and/or surface area, used evaluation tools that incorporated these measurements, or used complete wound healing as end points were included.

Individual trials used various terms to describe outcomes. Some trials used scales such as Pressure Ulcer Scale for Healing²² or Pressure Sore Status Tool.²³ For simplicity, remaining terms were classified into 3 categories: complete wound healing (ie, proportion of individuals whose wounds healed), time to healing (ie, time to complete wound healing), and wound surface area (ie, changes over time).

Methodological quality of the RCTs was determined using 6 elements from the checklist to evaluate a report of a nonpharmacological trial (CLEAR NPT) (http://www.bichat.inserm.fr/equipes/Emi0357/docs/usersguidelines.pdf)²⁴ that are relevant to therapies for pressure ulcers: (1) adequate allocation sequence generation (ie, use of an appropriate method to generate the randomization sequence); (2) concealed treatment allocation; (3) adequate participant blinding; (4) adequate outcome assessor blinding; (5) comparable rates of other treatments and care in each randomized group (eg, frequency of dressing changes); and (6) intention-to-treat analysis. If these elements were not explicitly reported, they were considered not performed. Three authors (M.R., S.R.K., W.W.) independently rated each RCT and reached consensus. Trials meeting 4 or more of the CLEAR NPT criteria were considered good quality. Trials meeting 3 or less of the CLEAR NPT criteria were considered suboptimal. We also assessed which articles reported a sample size justification to determine whether RCTs were adequately powered to detect either clinically important differences or equivalence of compared treatments.²⁵

Specialized support surfaces such as mattresses and cushions redistribute a patient's weight over skin and subcutaneous tissues as it presses against a bed or chair surface.²⁶ A reduction of pressure between the body and the support surface is considered helpful in healing pressure ulcers. The distinction between types of support surfaces is important because costs vary widely. Support surfaces were categorized using the National Pressure Ulcer Advisory Panel classification system²⁶: nonpowered (support surfaces such as foam that do not need electricity, previously known as static) and powered (support surfaces such as rotating beds that require electricity, previously known as dynamic). An overlay is a support surface designed to be placed on top of another support surface. Powered support surfaces are generally more expensive than nonpowered surfaces.⁶ Standard hospital mattresses (ie, not a specialized support surface) usually incur a 1-time cost of less than $200, but specialized support surfaces (frequently rented) can range from less than $5 per month for nonpowered mattress overlays to more than $3250 per month for some powered support surfaces.²⁷

Randomized controlled trials that described nutritional supplementation by any method (eg, enterally or parenterally) were included. Local wound care dressings were categorized by function rather than form (eg, films or gels).²⁸ Because many dressings perform more than 1 function, they were categorized based on their primary purpose: exudate absorbing (eg, foams), debriding (eg, collagenase), hydrating (eg, hydrocolloids), antimicrobial (eg, silver, povidone-iodine), and other (eg, did not fit any of these categories, fit in >1 category, or function was unclear). Adjunctive therapies were defined as modalities that neither directly address the underlying contributing factors nor primarily address local wound care (eg, vacuum therapy).

The search identified 872 abstracts, from which 103 relevant RCTs were selected. The flow diagram shows an overview of the study selection process (Figure). The 103 RCTs included 5889 participants. Only 15 trials involved more than 100 participants^29- 43 and 22 provided a sample size justification.^{29- 32,35,39- 40,42,44- 57}

Thirty-eight of the 103 trials took place in acute care (37%), 25 in mixed settings (24%), 22 in long-term care (21%), 6 in rehabilitation (6%), 4 in ambulatory care (4%), 3 in home care (3%), 1 in palliative care (1%), and 4 did not mention their treatment setting (4%). Twenty-two trials (21.4%) included only participants older than 60 years or described participants as elderly and 11 trials (10.7%) included only participants with spinal cord injuries.

Forty-five trials reported funding by the for-profit manufacturers of the products under evaluation (43.7%), 15 reported funding from nonprofit peer-reviewed granting agencies only (14.6%), 14 reported funding from for-profit and nonprofit organizations (13.6%), and 29 did not indicate sources of funding (28.2%). Eighty-three trials (80.6%) did not provide sufficient information about authors' potential financial conflicts of interest.

Three authors (M.R., S.K., W.W.) independently rated each RCT on CLEAR NPT items. Initial agreement was 83% (92% for adequate description of generation of allocation sequences, 81% for treatment allocation concealment, 82% for adequate participant blinding, 87% for adequate blinding of outcome assessors, 68% for co-interventions same in each group, and 90% for intention-to-treat analysis). Differences were resolved by consensus. Sixteen of the 103 trials (15.5%) met 4 or more of the CLEAR NPT criteria.

Nineteen RCTs (1572 participants) evaluated interventions for underlying contributing factors. Twelve RCTs (1214 participants) evaluated support surfaces (Table 1).^{29- 30,39- 42,44- 46,58- 60} None evaluated the effects of repositioning alone.

Quiz Ref IDSix RCTs directly compared powered (eg, alternating pressure) with nonpowered (eg, foam) support surfaces.^{30,41- 42,45,58- 59} Russell et al³⁰ and Day and Leonard⁴¹ found no differences in pressure ulcer healing between powered and nonpowered support surfaces. The remaining 4 studies, which met fewer quality criteria than Russell et al,³⁰ suggested that powered support surfaces were superior to nonpowered support surfaces,^{42,45,58- 59} although 1 RCT did not report statistical significance.⁵⁸ Inconsistent findings in these 6 studies result in persistent uncertainty regarding the benefit of powered support surfaces.

Five RCTs compared different types of powered mattresses. Evans et al⁶⁰ and Nixon et al⁴⁰ found no differences in ulcer healing between the 2 powered support surfaces they compared. Allman et al⁴⁴ found that ulcer surface area decreased with an air-fluidized mattress but increased on an alternating pressure mattress (median changes, −1.2 vs 0.5 cm² [95% confidence interval for the difference, −9.2 to −0.6 cm²]; P = .01).

Seven RCTs (358 participants) evaluated nutritional supplements (Table 2).^61- 67 All were oral supplements, but contents varied in each trial. Lee et al⁶¹ evaluated ulcer healing over 8 weeks in long-term care residents randomized to either a collagen protein supplement or placebo combined with standard care. Healing was measured with the Pressure Ulcer Scale for Healing (0 = healed, 17 = worst possible score).¹⁸ Individuals randomized to the supplement had better healing than those randomized to placebo (mean [SD] improvement in Pressure Ulcer Scale for Healing score, 3.55 [4.66] vs 3.22 [4.11], respectively; P < .05).

ter Riet et al⁶⁴ compared high-dose (500 mg twice daily) with low-dose (10 mg twice daily) vitamin C given for either 12 weeks or until pressure ulcer healing (whichever came first) and found no differences in wound closure rates or mean change in ulcer surface area per week. In contrast, Taylor et al⁶⁶ found that 500 mg of vitamin C twice daily was better than placebo (mean reduction in pressure ulcer area, 84% vs 42.7%; P < .005). Several factors may explain the disparate findings. The study by ter Riet et al,⁶⁴ but not the study by Taylor et al,⁶⁶ provided an adequate description of generation of allocation sequences and used an intention-to-treat analysis. In addition, the study by ter Riet et al⁶⁴ included more patients (88 vs 20 in Taylor et al⁶⁶), was multicentered, and had longer follow-up (12 vs 4 weeks, respectively). Thus, the value of vitamin C supplementation in pressure ulcer treatment remains uncertain.

Two other trials that found beneficial effects for nutritional supplements had suboptimal quality.^62- 63 One of these did not report statistical significance.⁶³ Desneves et al⁶² compared 3 diets in a study of 16 patients: standard hospital diet, standard hospital diet plus high protein, and standard hospital diet plus high protein with arginine, zinc, and antioxidants. The first 2 groups did not achieve significant improvements in pressure ulcer healing as measured by the Pressure Ulcer Scale for Healing, but the third group did.

Eleven of 19 studies (57.9%) evaluating underlying contributing factors adequately described the generation of random allocation sequences. Of the 19 studies, 8 indicated that participants were randomized using concealed allocation (42.1%).

Because blinding may be difficult when studying support surfaces, ratings for the CLEAR NPT item regarding participant blinding were not included in Table 1. It is feasible, however, to blind participants in nutritional supplement trials but this was done in only 4 of 7 trials. In all 19 RCTs evaluating underlying contributing factors, it was feasible to perform blinded outcome assessments, and this was described in 11 trials (57.9%).

Co-interventions were described as consistent in all treatment groups among 14 of the 19 studies (73.7%). Intention-to-treat analyses were described in only 3 of these 19 studies (15.8%). One support surface study⁶⁰ met all 5 CLEAR NPT criteria, 1 nutrition study⁶⁴ met 5 of 6 criteria, and two^30,61 of the 19 RCTs met 4 criteria.

Sixty-three RCTs (3330 participants) evaluated interventions targeting local wound care. Fifty-four RCTs (2857 participants) evaluated wound dressings (Table 3A and Table 3B).^{31- 32,34- 38,48- 53,55- 57,68- 105}Quiz Ref IDFive of the 7 highest-quality RCTs of wound dressings found no difference in wound healing with the products they compared: collagenase vs fibrinolysin or deoxyribonuclease, collagenase vs hydrocolloid, radiant heat dressing vs hydrocolloid and/or alginate and phenytoin solution vs normal saline.^{31,50,56,69,97} Sayag et al⁴⁹ performed a multicentered trial of 92 patients aged 60 years or older with pressure ulcers in acute care. They found that mean wound surface area reduction per week was 2.39 cm² (SD, 3.54) in wounds treated with calcium alginate and 0.27 cm² (SD, 3.21) in wounds treated with dextranomer paste (P<.001). Gerding and Browning⁹⁴ found oxyquinoline improved wound healing compared with lanolin or petrolatum. However, lanolin may cause allergic contact dermatitis and has fallen out of favor in chronic wound treatment.^106- 107 No debriding agent was consistently superior to other dressings for wound healing.^{31,48,68- 69,71}

Nine RCTs (473 participants) evaluated biological agents (Table 4).^33,108- 115 Three trials examined the effects of platelet-derived growth factors. The trial that met the most CLEAR NPT criteria was performed by Rees et al,³³ which compared 3 doses of recombinant human platelet–derived growth factor with placebo. The incidence of complete healing was greater in all 3 recombinant human platelet–derived growth factor groups (P < .03 in all groups) compared with placebo.

In another trial, nerve growth factor improved healing when compared with placebo at 6-week follow-up (mean [SD] reduction in pressure ulcer area, 738 [393] vs 485 [384] mm²; P = .03).¹¹²

Of the 63 studies examining local wound care, 22 adequately described the generation of random allocation sequences (34.9%) and 13 reported that participants were randomized using concealed allocation (20.6%). Only 15 of the 63 studies (23.8%) described adequate participant blinding. Adequate blinding of outcome assessors was described in 23 studies (36.5%). Co-interventions were equally applied in 28 studies (44.4%), and intention-to-treat analyses were performed in only 10 studies (15.9%). None of the 63 studies examining local wound care fulfilled all 6 CLEAR NPT criteria. One study⁴⁹ of dressings met 5 of the 6 criteria, 6 studies^{31,50,56,69,94,97} of dressings met 4 of the 6 criteria, and 2 studies^33,112 of biological agents met 4 of the 6 criteria. Fourteen of the 63 RCTs (22.2%) did not meet any of the CLEAR NPT criteria.^{32,48,55,74,81,84,86- 88,91,93,95,108}

Twenty-one RCTs (987 participants) evaluated adjunctive therapies (Table 5).^{43,47,54,116- 133}Quiz Ref IDAmong the good-quality RCTs examining adjunctive therapies, there were no benefits to the interventions, which included electric current (vs placebo electric current),⁵⁴ laser¹²⁷ (vs moist saline gauze), and ultrasound¹²³ (vs placebo ultrasound).

Two RCTs examined electromagnetic therapy and found improvements in wound healing compared with placebo or standard care,^125- 126 but 1 of these RCTs did not report statistical significance.¹²⁶ Four trials examined light therapy,^43,128- 130 with the highest-quality trial⁴³ demonstrating no improvement in healing with light therapy compared with placebo therapy.

Two RCTs studied vacuum therapy and found no improvement in wound healing compared with cadexomer iodine, papain-urea-chlorophyllin copper, or moist gauze.^116- 117 Of the 21 studies evaluating adjunctive therapies, 5 adequately described the generation of random allocation sequences (23.8%) and 2 provided information indicating that participants were randomized with concealed allocation (9.5%). Thirteen of the studies (61.9%) blinded participants adequately. Adequate blinding of outcome assessors was described in 14 of the 21 studies (66.7%).

Co-interventions were balanced between groups in 16 of the 21 studies (76.2%). Intention-to-treat analyses were performed in only 2 studies (9.5%). None of the 21 studies examining local wound care fulfilled all 6 criteria from the CLEAR NPT checklist. Two studies^54,127 of adjunctive therapies met 5 of the 6 criteria and 1 study¹²³ met 4 of 6 CLEAR NPT criteria.

Fundamental to chronic wound care are managing the underlying contributing factors, local wound care, and adjunctive therapies. Guidelines for the practical management of pressure ulcers³ are available from the Wound Healing Society (http://www3.interscience.wiley.com/journal/118605275/issue). Management of underlying contributing factors is likely more valuable in treating pressure ulcers than either topical or adjunctive therapies. Thus, priority should be given to addressing underlying causes.¹⁷ However, only 19 of 103 studies focused on management of underlying contributing factors, while the remaining 84 trials examined local wound care and adjunctive therapies. Overall, few RCTs demonstrated meaningful outcome differences between specific treatment strategies.

We did not find evidence that powered mattresses were superior to nonpowered mattresses. Quiz Ref IDSupport surfaces only address 1 aspect of pressure ulcer formation (ie, pressure), and not other important forces associated with immobility and ulcer formation (such as shear, friction, temperature, and moisture). To address the forces that contribute to ulcer formation, regular turning and transferring schedules may provide a less expensive alternative to costly support surfaces.¹³⁶ No trial examined optimal turning or transferring regimens.

We found little evidence that nutritional supplements improve pressure ulcer healing in patients without specific nutritional deficiencies. Protein supplementation of long-term care residents may be beneficial. None of the included RCTs documented nutritional deficiencies prior to nutrient supplementation, so it is uncertain whether the benefits of protein supplementation are limited to individuals who have protein deficiencies.

No single dressing was consistently superior to other dressings in the trials of pressure ulcers we examined. Similar results exist for other chronic wounds. Cochrane reviews have concluded that there is insufficient evidence to show any 1 dressing type better than others for arterial ulcers,⁸ venous stasis ulcers,⁹ or surgical wounds healing by secondary intention.¹¹ Standard local wound care for a healable pressure ulcer (ie, 1 with reversible underlying factors) should satisfy the 3 criteria of moisture balance, bacterial balance, and debridement.

Standard local wound care for a maintenance or nonhealable pressure ulcer may require antiseptics.^17,134- 135 Controversy persists in the literature regarding the efficacy and safety of antiseptics (such as povidone-iodine solution).^134,137 Two of the RCTs we examined compared antiseptics with moist dressings.^81,93 Neither of these trials met any CLEAR NPT criteria. Antiseptics are inexpensive and non-RCT evidence supports their continued use in maintenance or nonhealable wounds to help prevent wound deterioration.¹³⁴ Because no single dressing was superior to others, clinicians should select dressings that fulfill criteria for standard local wound care, while considering cost, ease of use, goals of care, and patient comfort.

Our results suggest recombinant human platelet–derived growth factor and nerve growth factor may improve healing, but further study is needed to confirm that these expensive agents provide value over standard care in clinical practice.^33,112

We found no evidence that adjunctive therapies improve pressure ulcer healing. A recent systematic review of vacuum therapy concluded that there is insufficient evidence to demonstrate clinical benefit, and the large number of prematurely terminated and unpublished trials of vacuum therapy is concerning.¹³⁸ No RCTs of hyperbaric oxygen therapy met our inclusion criteria. Two recent systematic reviews could not conclude if there was any benefit of hyperbaric oxygen therapy on pressure ulcers.^139- 140 Another systematic review¹⁴¹ found insufficient evidence to reach conclusions regarding the contributions of laser therapy, therapeutic ultrasound, electrotherapy, and electromagnetic therapy to chronic wound healing. Overall, there are limited data to support routine use of these expensive adjunctive therapies in managing pressure ulcers.

The methodological quality of the RCTs in our review was often inadequate. Only 1 of the 103 RCTs⁶⁰ met all of the quality standards we selected from the CLEAR NPT checklist. This may partly reflect the evolving understanding of how best to design and report RCTs evaluating nonpharmacological interventions.²⁴ The RCTs published after 1992 met many of the CLEAR NPT quality criteria.^{30- 31,33,38,49- 50,54,56,61,64,94,97,112,123,127} Only 22 of the 103 RCTs provided a sample size justification.^{29- 32,35,39- 40,42,44- 57} Many negative trials were likely underpowered to detect either clinically important differences or equivalence of the treatments they compared.²⁵

The paucity of high-quality RCTs evaluating pressure ulcer may reflect differences between regulatory requirements for medications vs pressure ulcer treatments such as dressings. Prescription medications must have demonstrated efficacy and safety in RCTs prior to attaining approval for marketing. In contrast, since passage of the 1997 Food and Drug Administration Modernization Act, dressing manufacturers are not required to submit evidence of safety or effectiveness to the US Food and Drug Administration before marketing a new product.¹⁴² Similar regulations are in place in other countries.¹⁴³ This situation raises concerns analogous to those highlighted by the lack of regulation for vitamins and herbal supplements since passage of the 1994 Dietary Supplement Health and Education Act.¹⁴⁴

Future RCTs will need to address the methodological deficiencies highlighted in this review. Studies also are needed to develop standardized methods for measuring wounds and reporting healing rates.¹⁹ Pressure ulcers may be too complex to successfully treat using a single modality.¹⁴⁵ Trials of multifactorial wound care interventions (eg, a combination of repositioning and local wound care) should be considered to determine whether they offer advantages over simpler interventions.

Our review has limitations. First, it was restricted to RCTs because they provide the best evidence of treatment efficacy. This is especially important given the multifaceted nature of pressure ulcer treatments and the importance of controlling for co-interventions. Nonetheless, evidence from nonrandomized trials also may provide insights into treatment benefits and risks. Second, we also restricted our review to trials published in the English language. Our examination of non-English trials suggests that including these trials would not have altered our results.

Third, we examined RCTs in a variety of settings. Results of some RCTs may not be generalizable to other populations. Comparing trials was complicated by the fact that different staging systems were used to categorize pressure ulcer severity. Finally, we likely underestimated information about potential conflicts of interest because many journals only recently began publishing this information.

Relatively few RCTs evaluating pressure ulcer treatments follow standard criteria for reporting nonpharmacological interventions. High-quality studies are needed to establish the efficacy and safety of many commonly used treatments. There is little evidence from RCTs to justify the use of 1 support surface or dressing over alternatives. Similarly, there is little evidence to justify the routine use of nutritional supplements, biological agents, and adjunctive therapies compared with standard care. Clinicians should make decisions regarding pressure ulcer therapy based on fundamental wound care principles, cost, ease of use, and patient preference.