Predictive Power of the Braden Scale for Pressure Sore Risk in Adult Critical Care Patients

Transcription

1 J Wound Ostomy Continene Nurs. 2012;39(6): Published by Lippinott Williams & Wilkins CE WOUND CARE Preditive Power of the Braden Sale for Pressure Sore Risk in Adult Critial Care Patients A Comprehensive Review Jill Cox Critial are is designed for managing the sikest patients within our healthare system. Multiple fators assoiated with an inreased likelihood of pressure uler development have been investigated in the ritial are population. Nevertheless, there is a lak of onsensus regarding whih of these fators poses the greatest risk for pressure uleration. While the Braden Sale for Pressure Sore Risk is the most ommonly used tool for measuring pressure uler risk in the United States, researh fousing on the umulative Braden Sale sore and subsale sores is laking in the ritial are population. This author onduted a literature review on pressure uler risk assessment in the ritial are population, to inlude the preditive value of both the total sore and the subsale sores. In this review, the subsales Sensory Pereption, Mobility, Moisture, and Frition/ Shear were found to be assoiated with an inreased likelihood of pressure uler development; in ontrast, the Ativity and Nutrition subsales were not found to predit pressure uler development in this population. In order to more preisely quantify risk in the ritially ill population, modifiation of the Braden Sale or development of a ritial are speifi risk assessment tool may be indiated. Introdution Pressure ulers (PUs) are enountered in all are settings, inluding the intensive are unit (ICU), and are desribed as perhaps the most underrated ondition within this are setting. 1 Despite implementation of evidene-based preventive interventions, hospital-aquired PUs ontinue to be a major healthare onern. In 2008, the Health Care Cost and Utilization Projet ited an 80% inrease in PU ourrene between the years 1993 and 2006 in hospitalized adult patients, with total assoiated osts estimated at $11 billion (US dollars). 2 From 2008 to 2009, there was a slight derease in the overall prevalene of hospitalaquired PUs. Nevertheless, prevalene rates in the ICU remained the highest among hospitalized patients, ranging from 9% to 42%. 3-5 In 2009, 3.3% of ritial are patients developed deep tissue injuries or stage III, IV, or unstageable ulers. 5 Pressure uler prevention has long been a major fous of patient are. Reent hanges enated by the US Centers for Mediare & Mediaid Servies restriting reimbursement for hospital-aquired stage III and IV (full-thikness) PUs have heightened awareness and inspired a renewed sense of urgeny for suessful PU-prevention programs in the aute and ritial are settings. 6 Despite quality are and best praties, PUs ontinue to develop in hospitalized patients and the risk is highest among those admitted to an ICU. 7-9 The first step in preventing PUs is aurate identifiation of patients at risk. Traditionally, PU risk measurement has been aomplished through the use of validated PU risk assessment tools suh as the Braden Sale for Pressure Sore Risk. 10 In the United States, the Braden Sale is the most widely used risk assessment tool aross all are settings, and it is reommended for use in multiple urrent linial pratie guidelines. 11, 12 While limited evidene suggests that the umulative Braden Sale sore predits PU risk in ritially ill patients, evidene onerning the ontribution of the instrument's 6 subsales is espeially sparse; only 4 studies were identified that examined the relationship of subsale sores to PU risk in this population In addition to the fators assessed via the Braden Sale, a number of other fators prevalent in ritially patients have been found to be assoiated with PU development. Jill Cox, PhD, RN, APN,C, CWOCN, Medial/Surgial Advaned Pratie Nurse/Wound, Ostomy and Continene Nurse, Englewood Hospital and Medial Center, Englewood, New Jersey and Assistant Professor of Nursing, Rutgers, The State University of New Jersey. The author delares no onflit of interest. Correspondene: Jill Cox, PhD, RN, APN,C, CWOCN, Englewood Hospital and Medial Center, 350 Engle St, Englewood, NJ ( jill.ox@ehm.om ). DOI: /WON.0b013e31826a4d83 Copyright 2012 by the Wound, Ostomy and Continene Nurses Soiety J WOCN November/Deember

2 614 Cox J WOCN November/Deember 2012 They inlude advaned age, 1, 4, 13, 14, lengths of stay in the ICU of greater than 5 days, 1,13,14,18,20 emergent admission to the ICU, 18, 21 severity of illness measured via the APACHE II sale, 1, 14, 22 and various omorbid onditions inluding diabetes mellitus, infetion, and ardiovasular/vasular disease. 13,14,17,19,23 Researhers have also evaluated the influene of iatrogeni fators on PU risk suh as the use of vasopressor agents. 1, 14, 23 At present, there is insuffiient data to determine the level of risk assoiated with these fators. Methods In order to gain a better understanding of the preditive power of the Braden Sale for Pressure Sore Risk in the ICU population, a omprehensive review of the literature was undertaken fousing on the preditive value of the overall Braden Sale sore and the individual subsale sores in determining PU risk in the ritial are population. The omputerized databases of EBSCO-CINAHL and EBSCO- MEDLINE were searhed using the terms pressure uler, Braden Sale, ritial are, intensive are and risk fators. Journal hand searhing and anestry searhing were also used as searh tehniques. Inlusion riteria established for this review inluded (1) peer reviewed and published reports on PU risk fators in adult ritial are patients that inluded the Braden Sale and/or subsales as variables and (2) studies onduted from 1995 to present. Exlusion riteria were (1) studies in languages other than English and (2) studies in whih interventions for PU prevention in ICU patients were the primary fous. Nine studies were identified that satisfied inlusion/ exlusion riteria. 4, 13-16, 19, Critial are settings represented in these studies inluded medial, surgial, medial/surgial, and neurologi ICUs. A summary of the studies inluded in this review an be found in Table 1. Braden Sale An Overview The Braden Sale for Prediting Pressure Sore Risk 10 measures umulative risk for PU development based on 7 risk fators measured on 6 subsales (Sensory Pereption, Ativity, Mobility, Moisture, Nutrition, and Frition/ Shear) and is based on the oneptual shema developed by Braden and Bergstrom. 27 Subsale sores range from 1 to 4 with the exeption of the Frition/Shear subsale, whih ranges from 1 to 3. Eah subsale sore is learly defined by narrative desriptors that assist the liniian to aurately math the patient's status to the orret subsale level. Pressure uler risk is based on a summated sore of 6 to 23, with lower sores indiating greater risk. Currently, a utoff sore of 18 has been found to demonstrate the best balane between sensitivity and speifiity; thus, linially this sore represents risk for PU development. 28 Some liniians propose stratifiation of PU risk, with sores of 15 to 18 indiating mild risk, sores of 13 to 14 indiating moderate risk, sores of 10 to 12 indiating high risk, and sores of 9 or less indiating very high risk. 29 The Braden Sale has been subjet to the most extensive psyhometri testing of all the PU risk tools. 30 Initial reliability studies, onduted in skilled nursing failities, yielded interrater reliability oeffiients ranging from r 0.83 to r 0.99 ( P.001). In the ritial are population, Bergstrom and olleagues 31 reported an interrater reliability of r 0.89 ( P.001). The preditive validity of a PU risk assessment tool an be assessed using 4 measurements: sensitivity, speifiity, positive preditive value (PPV), and negative preditive value (NPV) ( Figure 1 ). Sensitivity refers to the tool's ability to aurately identify patients at risk, and speifiity refers to the tool's ability to orretly identify patients not at risk, while PPV and NPV desribe the sale's ability to aurately predit patients who will and will not develop PUs. 30 In the ritial are population, preditive validity was first established by Bergstrom and olleagues, 31 who reported 83% sensitivity and 64% speifiity, with a NPV of 85% and a PPV of 61%, based on a utoff sore of 16. The Braden Sale has been reported to have performed similarly or better than other risk assessment instruments in various are settings on measures of preditive validity. 30, 32 Cumulative Risk A number of studies have used multivariate analyses to evaluate the Braden Sale in the ritial are setting. Total Braden Sale sore was a signifiant preditor of PU development in ritial are patients in 5 of the 9 studies reviewed. 4,13,15,19,24 In 1 study of 347 medial-surgial ICU patients, 14 statistially signifiant lower Braden Sale sores were found between patients who developed a PU and patients who did not develop a PU (Braden Sale sore 13 vs 15, respetively). In another study of 85 medialsurgial ICU patients, 16 the investigators reported a statistially signifiant differene in Braden Sale sores between patients who did and did not develop PUs; however, speifi sores were not reported. Wolverton and olleagues 26 found lower Braden Sale sores in their sample of 422 ritial are patients who developed PUs; however, their analysis was limited to desriptive statistis. In ontrast, a study of 40 medial ICU patients 25 found no signifiant differene in Braden Sale sores between patients who did and did not develop PUs. Cumulative sores on the Braden Sale in ritially ill patients varied from 9.8 to 15, indiating moderate to high risk. 13,14,15,19,25 The timing of the Braden Sale assessment varied aross the studies. In 3 studies, the Braden Sale sore was obtained at the time of admission to the ICU. 14, 16, 24 The timing of the sores in other studies inluded ross-setional measurements at the time of data olletion, 4, 13 mean Braden Sale sores reorded multiple

3 J WOCN Volume 39/Number 6 Cox 615 TABLE 1. Summary of Critial Care Studies Inluding the Braden Sale/Subsales as Variables Under Investigation (1995 to Present) (Chronologi Order) Study Authors Design Jirika and Exploratory olleagues 16 Desriptive Carlson and Prospetive design olleagues 15 with repeated measures Bours and Cross-setional olleagues 13 design seondary analyses Sample Size/ Type of ICU 85 surgial ICU 136 medial ICU 850 patients- Duth ICUs Fife and Prospetive ohort 186 neuro olleagues 24 ICU Pender and Desriptive Frazier 25 orrelational design Wolverton and olleagues medial ICU Desriptive study 422 medial/ surgial and neuro ICUs Shahin and Cross-setionalolleagues 4 Prevalene Slowikowsi Prospetive / and Funk 19 desriptive orrelational design Cox 14 Desriptive/ Correlational designretrospetive analysis 1760 medial, surgial ICU 369 surgial ICU 347 Medial/ Surgial ICU Pressure Uler Inidene, % Total Braden Sale Sore 56 b 12 Sensory Pereption Ativity Mobility Moisture 28.7 Nutrition Frition/ Shear None Other Risk Fators Signifiant in Multivariate Analysis a b DUPA sale: Moisture irulation Age, longer length of stay, infetion 12.4 Low BMI on admission 20 No multivariate analysis undertaken 13.7 No multivariate analysis undertaken 30 ( ); 16.2 (2006) Age, bowel inontinene 23.9 Age 70 y, diabetes mellitus 18.7 b Abbreviations: BMI, body mass index; DUPA, Deubitus Uler Potential Analyzer; ICU, intensive are unit. a Not all risk fators investigated are listed, only those fators found signifiant in multivariate analysis. b Signifiant in bivariate analysis only. Signifiant in multivariate analysis. b b Age, longer lengths of ICU stay, ardiovasular disease, norepinephrine infusion

4 616 Cox J WOCN November/Deember 2012 FIGURE 1. The elements of preditive validity for a pressure uler risk assessment sale. Based on data from Bolton. 30 times, 15 and lowest Braden Sale sore reorded during the ICU admission. 25 One study did not report when the Braden Sale sore was alulated. 19 In the studies reviewed, only 1 group of investigators provided data regarding interrater reliability. Jirika and oworkers 16 reported IRR measurements ranging from 88% to 92%; their study involved 85 medial-surgial ICU patients. Unfortunately, they provided neither any explanation for the range in IRR nor any desription as to how these data were obtained. No other studies reported measures of reliability in the published reports. Measurements of preditive validity, inluding sensitivity, speifiity, and NPVs and PPVs, were reported in 2 of the 9 studies in this review. 14, 16 Table 2 provides a summary of studies reporting the preditive validity of the Braden Sale in adult ritial are patients and ontains the 2 studies inluded in this review, 14, 16 the initial study on the preditive validity of the Braden Sale in the ICU population by Bergstrom and olleagues, 31 and 2 additional studies 33, 34 that foused exlusively on psyhometri testing of PU risk assessment sales in the ritial are population. While the majority of patients aross all study samples fell below the established level of 18 and were, therefore, onsidered at risk, the vast majority of patients remained PU free. In one study of 369 surgial ICU patients, % of the sample fell into the at-risk range (Braden Sale sore 18), with 60% of the sample reporting as high risk or very high risk. However, the inidene of PU development was only 24%. In a seond study of 347 medial/surgial ICU patients, 14 94% of the patients were lassified at risk, with 44% of the sample found to be at high risk or very high risk, but the reported inidene was 18.7%. Wolverton and assoiates 26 evaluated 422 patients in a medial/surgial ICU; they identified 92% as being at risk for PU, inluding 41% at high or very high risk, but the reported inidene of PU was 14%. This overpredition may represent a flaw in the risk assessment tool or it may reflet the positive effets of PU-prevention measures; this issue will be addressed in greater detail. Braden Subsales Four studies evaluated Braden subsales ( Table 1 ). Sensory pereption is defined as the ability of the individual to pereive and respond to disomfort as a result of exposure to pressure. 10 Examination of evidene reveals variability in reported influene of this subsale on PU development. Two studies 15, 16 reported that the Sensory Pereption subsale was a signifiant preditor of PU development. In a third study, 14 the Sensory Pereption subsale was signifiantly assoiated with PU development in a bivariate analysis; however, this subsale did not emerge as a signifiant preditor in a multivariate analysis. In ontrast, a fourth study found no statistially signifiant TABLE 2. Preditive Validity of the Braden Sale in Critial Care Studies Study Authors Cutoff sore Sensitivity, % Speifiity, % PPV, % NPV, % Braden and olleagues Jirika and olleagues Jirika and olleagues Seongsook and olleagues Cho and Noh Cox Abbreviations: NPV, negative preditive value; PPV, positive preditive value.

5 J WOCN Volume 39/Number 6 Cox 617 relationship between the Sensory Pereption subsale and PU development. 13 Braden and Bergstrom 27 define altered mobility as a diminished ability to hange and ontrol body position, whih inreases the potential for exposure to prolonged and intense pressure. Two studies found the Mobility subsale sore to be preditive of PU development. 13, 14 Conversely, 2 studies did not find this subsale to be signifiantly assoiated with PU development in either bivariate or multivariate analyses. 15, 16 Diminished levels of ativity (bed-bound or hairbound status) influene the duration and intensity of pressure experiened by patients and an ontribute to pressure uleration. 27 None of the 4 studies reporting Braden subsale sores found the Ativity subsale to be assoiated with PU development. Braden and Bergstrom 10 state that inreased maeration of the skin due to exposure to urine, stool, wound, or fistula drainage inreases its suseptibility to pressure uleration. Two studies 13, 16 found the Moisture subsale to be preditive of PU development, and 2 studies 14, 15 found no assoiation between the Moisture subsale and PU development. The Nutrition subsale is intended to reflet the individual's nutritional intake. 27 Nutritional defiienies lead to hypoproteinemi states and protein-alorie malnutrition, whih an alter the ability of the skin to tolerate prolonged exposure to pressure and inrease the risk for pressure uleration. Cox 14 evaluated the ontributions of the Braden Sale and subsales in a study of PU risk fators in 347 medial-surgial ICU patients. While the Nutrition subsale was found to be signifiantly assoiated with PU development in a bivariate analysis, this subsale was not a signifiant preditor of PU development in a multivariate analysis in this sample of ICU patients. While this is the only ICU study to find a signifiant bivariate relationship between the Nutrition subsale and PU development, no studies examining the Braden subsales found this subsale to be a signifiant preditor of PU development in this population. The Frition/Shear subsale measures 2 oneptually distint yet interrelated risk fators. Braden and Bergstrom 27 define frition as the fore that results when 2 surfaes move aross eah other suh as ours from dragging a patient to hange position. Shear is defined as a fore reated by the interplay of gravity and frition, resulting in damage at the deeper fasial levels. 35 While Jirika and olleagues 16 found a signifiant assoiation between the frition/shear subsale and PU development, Cox 14 found this subsale to be a signifiant preditor of PU development in this population. Disussion Considered olletively, findings from these studies provide evidene that the umulative Braden Sale sore is a signifiant preditor of PU risk in the ritial are setting. The Braden Sale is assoiated with lower speifiity and PPV sores, indiating a tendeny to overpredit PU development. Analysis of ritially ill subjets reveals that virtually all had a Braden Sale sore of 18 or less, but the majority did not develop a PU. Overpredition of PU inidene is often ritiized as a limitation shared by all of the validated PU risk assessment instruments. 30 There are 2 potential explanations of the omparatively low positive preditive sores reported in the studies under review. Administration of the Braden Sale may have aurately identified patients at risk for PU, resulting in the aggressive implementation of preventive strategies. Alternatively, it might be that the Braden Sale failed to adequately differentiate risk magnitude, resulting in the implementation of unneessary and potentially ostly preventive interventions. Based on urrent evidene, it is not possible to determine whih of these explanations is most aurate. Thus, aution is reommended when onsidering a risk assessment sale's preditive ability, beause the prevention strategies triggered by identifiation of a patient at risk an substantially redue the risk of PU development. 36 Another fator that may have influened the preditive validity of the Braden Sale is the timing of assessments within the individual studies. Some studies based findings on a Braden Sale sore obtained at admission, while others based findings on a single measurement or multiple measurements obtained over time. The Braden Sale sore obtained on admission is ritial from a linial perspetive, beause it enables prompt identifiation of risk and early implementation of prevention strategies. Basing statistial analysis on a Braden Sale sore obtained at admission provides a onsistent point in time of measurement for all subjets in a given study; however, it reflets only 1 risk assessment during the ICU admission and does not apture variability in patient and risk status throughout the ICU stay. Basing results on a ross-setional approah reflets risk sores obtained at various points during patients ICU admission and results in a single measurement obtained at different points during subjets ICU ourse. Analysis based on a single sore measured on admission or in a ross-setional analysis does not take into aount the variability in the auity of illness experiened by ritially ill patients during an ICU admission, whih an impat PU risk. Using multiple measurements over time for a statistial analysis has the potential advantage of apturing flutuations and trends in patients linial ondition and subsequent hanges in PU risk status. This approah minimizes the potential for use of a single aberrant risk sore in data analysis that may our with other study designs. Due to the variability in approah to Braden Sale measurement in these studies, the ability to translate the findings into the linial arena is hampered. Perhaps future investigators ould use a multiple measurement approah; suh an approah may

6 618 Cox J WOCN November/Deember 2012 enhane our ability to determine the true preditive validity of the Braden Sale. Consistent appliation of this approah may also provide liniians with empirial evidene regarding the optimal point or points in time for PU risk assessment in the ICU population. Evidene regarding the ontributions of the individual subsales to PU inidene in the ritial are population is espeially sparse and study findings are mixed. In the studies retrieved in this review, 4 subsales (Sensory Pereption, Mobility, Moisture, and Frition/Shear) demonstrated varying degrees of preditive value based on a multivariate analysis, and 2 (ativity and nutrition) did not prove preditive in any study. Determining the relative ontributions of eah of the subsales to overall risk is signifiant beause it helps identify whih preventive interventions are most appropriate for an individual patient. In the urrent era of heightened fisal aountability, implementation of risk-appropriate preventive measures is indiated in order to ensure the best possible linial and eonomi outomes. The Braden Sale Sensory Pereption subsale was found to be a signifiant risk fator in 2 of the 4 studies that examined the preditive validity of this subsale. 15, 16 These findings are onsistent with urrent linial pratie guidelines, 12 whih emphasize the signifiane of sensory impairment and reommend that liniians implement preventive measures for patients who sore low on the Sensory subsale, even if the total sore does not indiate signifiant risk. The Ativity subsale did not emerge as a signifiant risk fator in any of the 4 studies that evaluated subsales, but 2 studies 13, 14 reported that the Mobility subsale was a signifiant preditor of PU development. Though learly assoiated with mobility, ativity is defined as the patient's overall level of physial ativity and ranges from bedbound to ambulatory. 10 Sine most ritial are patients are bed-bound, this subsale has limited disriminating value when applied to a ritially ill population. Mobility takes into aount the patient's ability to move independently while onfined to bed. Therefore, it may provide a more aurate representation of the ritially ill patient's ability to adjust his or her position to redistribute pressure and prevent ishemia and uleration. The Braden subsales of Mobility, Ativity, and Sensory Pereption represent related yet oneptually distint risk fators. 10, 12 In the linial setting, espeially in the ICU population, these 3 risk fators (altered mobility, diminished ativity, and impaired sensory pereption) frequently oexist; thus, it an be diffiult to determine the degree to whih eah individual fator ontributes to overall PU risk. The Moisture subsale was found to be a signifiant preditor of PU development in 2 studies that enrolled ritial are patients, 13, 16 although a third study found no assoiation. 14 Two studies onduted in the ritial are setting found that feal inontinene was an independent preditor of PU development. 1, 4 Exposure of the skin to urine and stool is linked to an inreased risk for inontinene-assoiated dermatitis, 37 and both may be managed by indwelling devies in the ICU. Indwelling urinary atheters are ommon among ritially ill patients, espeially sine they are also used to provide an aurate measurement of urine output. Indwelling feal ontainment devies were introdued in 2004 and have gained popularity in linial pratie among ritially ill patients, although they are less widely used than urinary atheters. Preliminary evidene suggests that seletive use of these devies, ombined with an evidene-based PU-prevention program, may derease PU inidene among ICU patients exposed to high levels of moisture from liquid stool. 38 The Nutrition subsale was not preditive of PU development in any study retrieved for this review. The Nutrition subsale measures the patient's usual food intake. Beause of the auity of their illness, most ICU patients are unable to provide a dietary history, whih limits the preditive value of this subsale in this population. Sparse evidene suggests that alternative tehniques for measuring the nutritional status of ritially ill patients may prove more preditive. For example, body mass index and the number of days without nutrition have been reported as signifiant preditors of PU inidene in the ICU population. 18, 24 At this point, we need more data in order to identify the best physiologi indiator of nutritional status in the ritially ill patient. 39 Commonly used biomarkers suh as body weight, albumin, prealbumin, and lymphoyte ounts are of limited value due to the intravasular fluid shifts that are ommon in ritial are patients and that signifiantly influene both weight and laboratory values. In addition, evidene linking nutritional status to PU risk is mixed, and additional researh is needed to more preisely define the nature of this relationship. 12, 40 The Frition/Shear subsale is inluded in the Braden Sale beause eah of these fores an ause signifiant damage to the skin and soft tissue. Frition damages the epidermal and dermal layers of the skin, while shearing fores ause angulation and deformation of the blood vessels at the fasia level. Shear fores have been hypothesized to ause muh of the damage assoiated with full-thikness pressure uleration. 35 The Frition/Shear subsale emerged as a signifiant preditor of PU development in 1 study. 14 Many ICU patients require prolonged head-of-bed elevation for prevention of ventilator-assoiated pneumonia and/or to prevent aspiration in patients reeiving enteral feedings; head-of-bed elevation predisposes the patient to signifiant shearing fores. Researh evaluating the effets of prolonged head-of-bed elevation on PU inidene is warranted in an effort to better understand the effets of shear fores on skin and tissue integrity. 41 The results of these studies suggest that the linial utility of the Braden Sale may be limited. Most ritially ill patients are deemed at risk for PU development when assessed using the Braden Sale, and it is not yet known whether the omparatively low preditive values represent

7 J WOCN Volume 39/Number 6 Cox 619 the suess of preventive interventions or failure of the tool to aurately differentiate between patients who are at risk and those who are not at risk. In addition, there are fators unique to the ritial are population that may inrease their risk for PU development, suh as the use of vasopressor agents, prolonged ICU admissions, and omorbid onditions; further investigation is needed to determine their ontribution to PU risk. If these fators are found to be preditive of PU development in this population, they should be inorporated into a setting-speifi, validated risk assessment tool. 1, 13, 14, 17-20, 23 Other fators suh as advaning age and low arterial pressure were hypothesized to be related to PU development by Braden and Bergstrom, even though they ultimately were not inluded in the Braden Sale for Pressure Sore Risk. 27 A growing body of evidene suggests that advaning age inreases the likelihood of PU development in the ritial are population, 1, 4, 13, 14, whereas studies to date have not shown low arterial pressure to be a risk fator for PU development in this population. 14,19,23,25,42 Impliations for Researh and Pratie The findings of this review reveal multiple opportunities for additional researh inluding development and testing of a PU risk assessment tool speifi to the ritial are population, either a modified Braden Sale or a newly developed tool. There is preedene for development of a modified Braden Sale; the Braden Q Risk Assessment Tool is a modified and validated version of the Braden Sale that inludes the 6 Braden subsales and an additional subsale measuring tissue perfusion and oxygenation; it is designed for use in the pediatri population. 43, 44 Modifiation of the urrent Braden Sale may help to more aurately identify ritial are patients who are at signifiant risk for PU development. Redefining the various levels within the subsale definitions so that they are relevant to the ritial are population may be one potential option. The inlusion of other empirially supported ritial are risk fators suh as advaned age, prolonged ICU length of stay, omorbid onditions, or vasopressor use should undergo further investigation to determine whether their inlusion might enhane the preditive power of suh an instrument in the ritial are setting. Pressure uler risk assessment sales, inluding the Braden Sale, tend to overpredit risk; as noted, this may be due to an inherent weakness in the tool itself or may reflet the effetiveness of urrently used prevention protools. 30 The majority of ICU patients in this review were found to be at risk for PU development based on the Braden Sale sore but did not develop a PU; it is unknown whether this represents true overpredition or is the result of preventive are. In the first senario, overpredition may be the result of an intrinsi weakness of the sale and results in the unneessary implementation of prevention protools, whih ould impat healthare osts. In this ase, the refinement or development of a sale that better measures PU risk in the population would be warranted. In the seond senario, the apparent overpredition may reflet the suessful implementation of PU-prevention protools; identifiation of the patient as being at risk triggered preventive are that atually prevented PU ourrene. Clinially, the seond senario validates the benefits of a omprehensive PU-prevention program. Sine withholding PU-prevention strategies would be unethial, it is impossible to ondut a study to definitively determine whether the apparent overpredition is true overpredition or the result of effetive are. In linial pratie, the onsequenes of underpredition would far outweigh the osts of overpredition. 30 An in-depth analysis of at-risk ritial are patients who develop a PU despite prevention strategies ompared to at-risk ritial are patients who do not develop a PU may be of benefit. Suh an analysis might provide insight into the risk fators that ontribute to PU development in this population that are not measured by the Braden Sale and ould provide empirial evidene for the development of a ritial are risk assessment sale or a modified Braden Sale. Suh a study might also provide valuable linial information regarding the effetiveness of the PU-prevention program. Development of a PU in a patient who has been identified as at risk and plaed on a prevention protool may represent some failure or gap in the PU-prevention protool or may represent an unavoidable uler. The purpose of all PU risk assessment sales is to predit PU risk and subsequently mobilize liniians to implement prevention strategies that will impede PU ourrene. Thus, the determination of the preditive validity of a PU risk assessment sale annot be made in isolation from the prevention strategies that are implemented. However, it is possible that a risk assessment tool designed speifially for the ritial are population ould help eliminate true overpredition and to more aurately identify patients who need preventive are; thus, further researh in this area is warranted. Conlusion Researh indiates that ritially ill patients who develop PUs are lassified as at risk by the Braden Sale, and that most of the ritial are patients who did not develop PUs were also lassified as at risk. At present, we do not know whether this disrepany reflets the suess of preventive interventions or inadequate disrimination of risk. While the subsales Sensory Pereption, Mobility, Frition/Shear, and Moisture have demonstrated preditive ability in ritial are patients, the pauity of empirial investigations preludes our ability to draw definitive onlusions regarding the relative ontributions of eah of these subsales to PU risk detetion and PU development. Modifiation of the Braden Sale or development of a ritial are PU risk assessment sale might improve our

8 620 Cox J WOCN November/Deember 2012 ability to aurately identify at-risk patients in this population and should be addressed in future studies. KEY POINTS Limited linial evidene suggests that the umulative Braden Sale sore is a preditor of PU development in adult ritial are patients. Evidene onerning the ontribution of Braden Sale subsales is sparse, and findings are mixed in the ritial are population. Limited evidene suggests that the Sensory Pereption, Mobility, Moisture, and Frition/Shear subsales are preditors of PU development in the ritial are population. Modifiation of the Braden Sale or development of a ritial are speifi risk assessment tool might improve our ability to aurately identify at-risk patients in the ritial are setting. Referenes 1. Theaker C, Mannan M, Ives N, Soni N. Risk fators for pressure sores in the ritially ill. Anaesthesia ; 55 : Russo C, Steiner C, Spetor W. Hospitalizations related to pressure ulers among adults 18 years and older Healthare Cost and Utilization Projet. Ageny for Healthare Quality and Researh Statistial Brief Published Aessed February 28, Keller P, Wille J, vanramshorst B, van der Werken C. Pressure ulers in intensive are patients: a review of risks and prevention. Intensive Care Med ; 28 : Shahin ES, Dassen T, Halfens RJ. Pressure uler prevalene in intensive are patients: a ross-setional study. J Eval Clin Pra ; 14 : Van Gilder C, Amlung S, Harrison P, Meyer S. Results of the International Pressure Uler Prevalene Survey and a 3-year aute are, unit-speifi analysis. Ostomy Wound Manage ; 55 ( 11 ): Centers for Mediare & Mediaid Servies. CMS proposes additions to hospital aquired onditions for fisal year nter 3042&intNumPerPage 10&hekDate &hekkey & srhtype 1&numDays 3500&srhOpt 0&srhData &key wordtype All&hkNewsType 6&intPage &showall &pyea r &year &des &boorder date. Published Aessed February 28, Baumgarten M, Margolis D, Berlin J, et al. Risk fators for pressure ulers among elderly hip frature patients. Wound Repair Regen ; 11 ( 2 ): Baumgarten M, Margolis D, Loalio R. Extrinsi risk fators for pressure ulers early in the hospital stay: a nested ase-ontrol study. J Gerontol A Biol Si Med Si ; 63A ( 4 ): Krapfl L, Makey D. Mediare hanges to the hospital inpatient prospetive payment systems. J Wound Ostomy Continene Nurs ; 35 ( 1 ): Bergstrom N, Braden B, Laguzza A, Holman V. The Braden Sale for prediting pressure sore risk. Nurs Res ; 36 ( 4 ): Wound, Ostomy and Continene Nurses Soiety. Guideline for the Prevention and Management of Pressure Ulers. Mount Laurel, NJ : Wound, Ostomy and Continene Nurses Soiety National Pressure Uler Advisory Panel and European Pressure Uler Advisory Panel. Prevention and Treatment of Pressure Ulers: Clinial Pratie Guidelines. Washington, DC : National Pressure Uler Advisory Panel ; Bours G, De Laat E, Halfens R, Lubbers M. Prevalene, risk fators and prevention of pressure ulers in Duth intensive are units. Intensive Care Med ; 27 : Cox J. Preditors of pressure uler development in adult ritial are patients. Am J Crit Care ; 20 ( 5 ): Carlson E, Kemp M, Shott S. Prediting the risk of pressure ulers in ritially ill patients. Am J Crit Care ; 8 ( 4 ): Jirika M, Ryan P, Carvalho M, Bukvih J. Pressure uler risk fators in an ICU population. Am J Crit Care ; 4 ( 5 ): Frankel H, Sperry J, Kaplan L. Risk fators for pressure uler development in a best pratie surgial intensive are unit. Am Surg ; 73 : Eahempati S, Hydo L, Barie P. Fators influening the development of deubitus ulers in ritially ill surgial patients. Crit Care Med ; 29 ( 9 ): Slowikowski G, Funk M. Fators assoiated with pressure ulers in patients in a surgial intensive are unit. J Wound Ostomy Continene Nurs ; 37 ( 6 ): Sayar S, Turgut S, Dogan H, et al. Inidene of pressure ulers in intensive are unit patients at risk aording to the Waterlow sale and fators influening the development of pressure ulers. J Clin Nurs ; Kaitani T, Tokunaga K, Matsui N, Sanada H. Risk fators related to the development of pressure ulers in the ritial are setting. J Clin Nurs ; 19 : Shahin E, Dassen T, Halfens R. Inidene, prevention and treatment of pressure ulers in intensive are patients: a longitudinal study. Int J Nurs Stud ; 46 : Batson S, Adam S, Hall G, Quirke S. The development of a pressure area soring system for ritially ill patients: a pilot study. Intensive Crit Care Nurs ; 9 : Fife C, Otto G, Capsuto E, et al. Inidene of pressure ulers in a neurologi intensive are unit. Crit Care Med ; 29 ( 2 ): Pender L, Frazier S. The relationship between dermal ulers, oxygenation and perfusion in mehanially ventilated patients. Intensive Crit Care Nurs ; 21 : Wolverton C, Hobbs L, Beeson T, et al. Nosoomial pressure uler rates in ritial are. J Nurs Care Qual ; 20 ( 1 ): Braden B, Bergstrom N. A oneptual shema for the study of the etiology of pressure sores. Rehabil Nurs ; 12 ( 1 ): 8-12, Bergstrom N, Braden B, Kemp M, Champagne M, Ruby E. Prediting pressure uler risk: a multisite study of the preditive validity of the Braden Sale. Nurs Res ; 47 ( 5 ): Prevention Plus. Home of the Braden Sale. Protools by level of risk. Published Aessed Otober 5, Bolton L. Whih pressure uler risk assessment sales are valid for linial use? J Wound Ostomy Continene Nurs ; 34 ( 4 ):

9 J WOCN Volume 39/Number 6 Cox Bergstrom N, Demuth P, Braden B. A linial trial of the Braden Sale for prediting pressure sore risk. Nurs Clin North Am ; 22 ( 2 ): Panorbo-Hildago T, Garia-Fernandez F, Lopez-Madina I, Alvarez-Nieto C. Risk assessment sales for pressure uler prevention: a systemati review. J Adv Nurs ; 54 ( 1 ): Seongsook J, Ihnsook J, Younghee L. Validity of pressure uler risk assessment sales: Cubbin and Jakson, Braden and Douglas sale. Int J Nurs Stud ; 41 ( 2 ): Cho I, Noh M. Braden Sale: evaluation of linial usefulness in an intensive are unit. J Adv Nurs ; 66 ( 2 ): Pieper B. Pressure ulers: impat, etiology and lassifiation. In: Bryant R, Nix D, eds. Aute and Chroni Wounds: Current Management Conepts. 4th ed. St Louis, MO : Mosby ; 2012 : Defloor T, Grypdonk M. Validation of pressure uler risk assessment sales: a ritique. J Adv Nurs ; 48 ( 6 ): Blak JM, Gray M, Bliss DZ, et al. MASD part 2: inontinene-assoiated dermatitis and intertriginous dermatitis: a onsensus. J Wound Ostomy Continene Nurs ; 38 ( 4 ): Benoit R, Watts C. The effet of a pressure uler prevention program and the bowel management system in reduing pressure uler prevalene in an ICU setting. J Wound Ostomy Continene Nurs ; 34 ( 2 ): Doughty D. Prevention and early detetion of pressure ulers in hospitalized patients: ommentary. J Wound Ostomy Continene Nurs ; 35 ( 1 ): Langer G, Knerr A, Kuss O, Behrens J, Shlomer GJ. Nutritional interventions for preventing and treating pressure ulers. Cohrane Database Syst Rev. 2008; ( 4 ): CD National Pressure Uler Advisory Panel. Shear: a ontributory fator in pressure uler development. National Pressure Uler Advisory Panel Shear fore initiative. npuap.org/shear_slides.pdf. Published Aessed Marh 6, Compton F, Hoffmann F, Hortig T, et al. Pressure uler preditors in ICU patients: nursing skin assessment versus objetive parameters. J Wound Care ; 17 ( 10 ): Curley M, Razmus I, Roberts K, Wypij D. Prediting pressure uler risk in pediatri patients. Nurs Res ; 52 ( 1 ): Quigley S, Curley M. Skin integrity in the pediatri population: preventing and managing pressure ulers. J So Pediatr Nurs ; 1 ( 1 ): For more than 70 additional ontinuing eduation artiles related to skin/wound are and 31 additional ontinuing eduation artiles related to researh, go to NursingCenter.om/CE