Professional Documents
Culture Documents
Predictive Power of The Braden Scale For Pressure Sore Risk in Adult Critical Care Patients
Predictive Power of The Braden Scale For Pressure Sore Risk in Adult Critical Care Patients
CE WOUND CARE
Critical care is designed for managing the sickest patients patients developed deep tissue injuries or stage III, IV, or
within our healthcare system. Multiple factors associated with unstageable ulcers.5
an increased likelihood of pressure ulcer development have Pressure ulcer prevention has long been a major focus
been investigated in the critical care population. Nevertheless, of patient care. Recent changes enacted by the US Centers
there is a lack of consensus regarding which of these factors for Medicare & Medicaid Services restricting reimburse-
poses the greatest risk for pressure ulceration. While the ment for hospital-acquired stage III and IV (full-thickness)
Braden Scale for Pressure Sore Risk is the most commonly used PUs have heightened awareness and inspired a renewed
tool for measuring pressure ulcer risk in the United States, re- sense of urgency for successful PU-prevention programs in
search focusing on the cumulative Braden Scale score and sub- the acute and critical care settings.6 Despite quality care
scale scores is lacking in the critical care population. This author and best practices, PUs continue to develop in hospital-
conducted a literature review on pressure ulcer risk assessment ized patients and the risk is highest among those admitted
in the critical care population, to include the predictive value of to an ICU.7-9
both the total score and the subscale scores. In this review, the The first step in preventing PUs is accurate identifica-
subscales Sensory Perception, Mobility, Moisture, and Friction/ tion of patients at risk. Traditionally, PU risk measurement
Shear were found to be associated with an increased likelihood has been accomplished through the use of validated PU
of pressure ulcer development; in contrast, the Activity and risk assessment tools such as the Braden Scale for Pressure
Nutrition subscales were not found to predict pressure ulcer de- Sore Risk.10 In the United States, the Braden Scale is the
velopment in this population. In order to more precisely quan- most widely used risk assessment tool across all care set-
tify risk in the critically ill population, modification of the tings, and it is recommended for use in multiple current
Braden Scale or development of a critical care specific risk as- clinical practice guidelines.11,12 While limited evidence
sessment tool may be indicated. suggests that the cumulative Braden Scale score predicts
PU risk in critically ill patients, evidence concerning the
contribution of the instrument's 6 subscales is especially
■ Introduction sparse; only 4 studies were identified that examined the
Pressure ulcers (PUs) are encountered in all care settings, relationship of subscale scores to PU risk in this
including the intensive care unit (ICU), and are described population.13-16
as perhaps the most underrated condition within this care In addition to the factors assessed via the Braden Scale,
setting.1 Despite implementation of evidence-based pre- a number of other factors prevalent in critically patients
ventive interventions, hospital-acquired PUs continue to have been found to be associated with PU development.
be a major healthcare concern. In 2008, the Health Care
Cost and Utilization Project cited an 80% increase in PU 䡲 Jill Cox, PhD, RN, APN,C, CWOCN, Medical/Surgical Advanced
occurrence between the years 1993 and 2006 in hospital- Practice Nurse/Wound, Ostomy and Continence Nurse, Englewood
Hospital and Medical Center, Englewood, New Jersey and Assistant
ized adult patients, with total associated costs estimated at
Professor of Nursing, Rutgers, The State University of New Jersey.
$11 billion (US dollars).2 From 2008 to 2009, there was a The author declares no conflict of interest.
slight decrease in the overall prevalence of hospital- Correspondence: Jill Cox, PhD, RN, APN,C, CWOCN, Englewood
acquired PUs. Nevertheless, prevalence rates in the ICU Hospital and Medical Center, 350 Engle St, Englewood, NJ 07631
remained the highest among hospitalized patients, rang- (jill.cox@ehmc.com).
ing from 9% to 42%.3-5 In 2009, 3.3% of critical care DOI: 10.1097/WON.0b013e31826a4d83
Copyright © 2012 by the Wound, Ostomy and Continence Nurses Society™ J WOCN ■ November/December 2012 613
Copyright © 2012 Wound, Ostomy and Continence Nurses Society™. Unauthorized reproduction of this article is prohibited.
Copyright © 2012 Wound, Ostomy and Continence Nurses Society™. Unauthorized reproduction of this article is prohibited.
WON200422.indd 615
TABLE 1.
Summary of Critical Care Studies Including the Braden Scale/Subscales as Variables Under Investigation (1995 to Present) (Chronologic Order)
Other Risk Factors
Study Sample Size/ Pressure Ulcer Total Braden Sensory Nutrit- Friction/ Significant in Multivariate
Authors Design Type of ICU Incidence, % Scale Score Perception Activity Mobility Moisture ion Shear Analysisa
■ Volume 39/Number 6
Copyright © 2012 Wound, Ostomy and Continence Nurses Society™. Unauthorized reproduction of this article is prohibited.
615
31/10/12 1:55 PM
616 Cox J WOCN ■ November/December 2012
FIGURE 1. The elements of predictive validity for a pressure ulcer risk assessment scale.
Based on data from Bolton.30
times,15 and lowest Braden Scale score recorded during the reporting as high risk or very high risk. However, the in-
ICU admission.25 One study did not report when the cidence of PU development was only 24%. In a second
Braden Scale score was calculated.19 study of 347 medical/surgical ICU patients,14 94% of the
In the studies reviewed, only 1 group of investigators patients were classified at risk, with 44% of the sample
provided data regarding interrater reliability. Jiricka and found to be at high risk or very high risk, but the reported
coworkers16 reported IRR measurements ranging from 88% incidence was 18.7%. Wolverton and associates26 evalu-
to 92%; their study involved 85 medical-surgical ICU ated 422 patients in a medical/surgical ICU; they identi-
patients. Unfortunately, they provided neither any expla- fied 92% as being at risk for PU, including 41% at high or
nation for the range in IRR nor any description as to how very high risk, but the reported incidence of PU was 14%.
these data were obtained. No other studies reported mea- This “overprediction” may represent a flaw in the risk as-
sures of reliability in the published reports. sessment tool or it may reflect the positive effects of
Measurements of predictive validity, including sensi- PU-prevention measures; this issue will be addressed in
tivity, specificity, and NPVs and PPVs, were reported in 2 greater detail.
of the 9 studies in this review.14,16 Table 2 provides a sum-
mary of studies reporting the predictive validity of the
Braden Scale in adult critical care patients and contains
■ Braden Subscales
the 2 studies included in this review,14,16 the initial study Four studies13-16 evaluated Braden subscales (Table 1).
on the predictive validity of the Braden Scale in the ICU Sensory perception is defined as the ability of the indi-
population by Bergstrom and colleagues,31 and 2 addi- vidual to perceive and respond to discomfort as a result of
tional studies33,34 that focused exclusively on psychometric exposure to pressure.10 Examination of evidence reveals
testing of PU risk assessment scales in the critical care variability in reported influence of this subscale on PU de-
population. velopment. Two studies15,16 reported that the Sensory
While the majority of patients across all study samples Perception subscale was a significant predictor of PU de-
fell below the established level of 18 and were, therefore, velopment. In a third study,14 the Sensory Perception sub-
considered “at risk,” the vast majority of patients re- scale was significantly associated with PU development in
mained PU free. In one study of 369 surgical ICU pa- a bivariate analysis; however, this subscale did not emerge
tients,19 99.5% of the sample fell into the at-risk range as a significant predictor in a multivariate analysis. In con-
(Braden Scale score ⱕ 18), with 60% of the sample trast, a fourth study found no statistically significant
TABLE 2.
Predictive Validity of the Braden Scale in Critical Care Studies
Study Authors Cutoff score Sensitivity, % Specificity, % PPV, % NPV, %
Braden and colleagues 27 16 83 64 61 85
Jiricka and colleagues 16 15 100 10.8 59.3 100
Jiricka and colleagues16 11 75 65 73.5 66.7
Seongsook and colleagues33 16 97 26 37 95
Cho and Noh 34 13 75.9 47.3 18.1 92.8
Cox 14 18 100 7 20 100
Abbreviations: NPV, negative predictive value; PPV, positive predictive value.
Copyright © 2012 Wound, Ostomy and Continence Nurses Society™. Unauthorized reproduction of this article is prohibited.
relationship between the Sensory Perception subscale and significant predictor of PU risk in the critical care setting.
PU development.13 The Braden Scale is associated with lower specificity and
Braden and Bergstrom27 define altered mobility as a di- PPV scores, indicating a tendency to overpredict PU devel-
minished ability to change and control body position, opment. Analysis of critically ill subjects reveals that virtu-
which increases the potential for exposure to prolonged ally all had a Braden Scale score of 18 or less, but the
and intense pressure. Two studies found the Mobility sub- majority did not develop a PU. Overprediction of PU inci-
scale score to be predictive of PU development.13,14 dence is often criticized as a limitation shared by all of the
Conversely, 2 studies did not find this subscale to be sig- validated PU risk assessment instruments.30
nificantly associated with PU development in either bi- There are 2 potential explanations of the compara-
variate or multivariate analyses.15,16 tively low positive predictive scores reported in the studies
Diminished levels of activity (bed-bound or chair- under review. Administration of the Braden Scale may
bound status) influence the duration and intensity of pres- have accurately identified patients at risk for PU, resulting
sure experienced by patients and can contribute to in the aggressive implementation of preventive strategies.
pressure ulceration.27 None of the 4 studies reporting Alternatively, it might be that the Braden Scale failed to
Braden subscale scores found the Activity subscale to be adequately differentiate risk magnitude, resulting in the
associated with PU development. implementation of unnecessary and potentially costly
Braden and Bergstrom10 state that increased macera- preventive interventions. Based on current evidence, it is
tion of the skin due to exposure to urine, stool, wound, or not possible to determine which of these explanations is
fistula drainage increases its susceptibility to pressure ul- most accurate. Thus, caution is recommended when con-
ceration. Two studies13,16 found the Moisture subscale to be sidering a risk assessment scale's predictive ability, because
predictive of PU development, and 2 studies14,15 found no the prevention strategies triggered by identification of a
association between the Moisture subscale and PU patient “at risk” can substantially reduce the risk of PU
development. development.36
The Nutrition subscale is intended to reflect the indi- Another factor that may have influenced the predic-
vidual's nutritional intake.27 Nutritional deficiencies lead tive validity of the Braden Scale is the timing of assess-
to hypoproteinemic states and protein-calorie malnutri- ments within the individual studies. Some studies based
tion, which can alter the ability of the skin to tolerate pro- findings on a Braden Scale score obtained at admission,
longed exposure to pressure and increase the risk for while others based findings on a single measurement or
pressure ulceration. Cox14 evaluated the contributions of multiple measurements obtained over time. The Braden
the Braden Scale and subscales in a study of PU risk factors Scale score obtained on admission is critical from a clinical
in 347 medical-surgical ICU patients. While the Nutrition perspective, because it enables prompt identification of
subscale was found to be significantly associated with PU risk and early implementation of prevention strategies.
development in a bivariate analysis, this subscale was not Basing statistical analysis on a Braden Scale score obtained
a significant predictor of PU development in a multivari- at admission provides a consistent point in time of mea-
ate analysis in this sample of ICU patients. While this is surement for all subjects in a given study; however, it re-
the only ICU study to find a significant bivariate relation- flects only 1 risk assessment during the ICU admission and
ship between the Nutrition subscale and PU development, does not capture variability in patient and risk status
no studies examining the Braden subscales found this sub- throughout the ICU stay. Basing results on a cross-sec-
scale to be a significant predictor of PU development in tional approach reflects risk scores obtained at various
this population. points during patients’ ICU admission and results in a
The Friction/Shear subscale measures 2 conceptually single measurement obtained at different points during
distinct yet interrelated risk factors. Braden and Bergstrom27 subjects’ ICU course. Analysis based on a single score mea-
define friction as the force that results when 2 surfaces sured on admission or in a cross-sectional analysis does
move across each other such as occurs from dragging a not take into account the variability in the acuity of illness
patient to change position. Shear is defined as a force cre- experienced by critically ill patients during an ICU admis-
ated by the interplay of gravity and friction, resulting in sion, which can impact PU risk. Using multiple measure-
damage at the deeper fascial levels.35 While Jiricka and col- ments over time for a statistical analysis has the potential
leagues16 found a significant association between the fric- advantage of capturing fluctuations and trends in patients’
tion/shear subscale and PU development, Cox14 found this clinical condition and subsequent changes in PU risk sta-
subscale to be a significant predictor of PU development tus. This approach minimizes the potential for use of a
in this population. single aberrant risk score in data analysis that may occur
with other study designs. Due to the variability in ap-
proach to Braden Scale measurement in these studies, the
■ Discussion ability to translate the findings into the clinical arena is
Considered collectively, findings from these studies pro- hampered. Perhaps future investigators could use a mul-
vide evidence that the cumulative Braden Scale score is a tiple measurement approach; such an approach may
Copyright © 2012 Wound, Ostomy and Continence Nurses Society™. Unauthorized reproduction of this article is prohibited.
Copyright © 2012 Wound, Ostomy and Continence Nurses Society™. Unauthorized reproduction of this article is prohibited.
the success of preventive interventions or failure of the case, the refinement or development of a scale that better
tool to accurately differentiate between patients who are measures PU risk in the population would be warranted.
at risk and those who are not at risk. In addition, there are In the second scenario, the apparent overprediction may
factors unique to the critical care population that may in- reflect the successful implementation of PU-prevention
crease their risk for PU development, such as the use of protocols; identification of the patient as being “at risk”
vasopressor agents, prolonged ICU admissions, and co- triggered preventive care that actually prevented PU oc-
morbid conditions; further investigation is needed to de- currence. Clinically, the second scenario validates the ben-
termine their contribution to PU risk. If these factors are efits of a comprehensive PU-prevention program. Since
found to be predictive of PU development in this popula- withholding PU-prevention strategies would be unethical,
tion, they should be incorporated into a setting-specific, it is impossible to conduct a study to definitively deter-
validated risk assessment tool.1,13,14,17-20,23 Other factors mine whether the apparent overprediction is true overpre-
such as advancing age and low arterial pressure were hy- diction or the result of effective care. In clinical practice,
pothesized to be related to PU development by Braden and the consequences of underprediction would far outweigh
Bergstrom, even though they ultimately were not included the costs of overprediction.30
in the Braden Scale for Pressure Sore Risk.27 A growing An in-depth analysis of “at-risk” critical care patients
body of evidence suggests that advancing age increases the who develop a PU despite prevention strategies compared
likelihood of PU development in the critical care popula- to “at-risk” critical care patients who do not develop a PU
tion,1,4,13,14,17-19 whereas studies to date have not shown low may be of benefit. Such an analysis might provide insight
arterial pressure to be a risk factor for PU development in into the risk factors that contribute to PU development in
this population.14,19,23,25,42 this population that are not measured by the Braden Scale
and could provide empirical evidence for the development
of a critical care risk assessment scale or a modified Braden
■ Implications for Research and Practice Scale. Such a study might also provide valuable clinical in-
The findings of this review reveal multiple opportunities formation regarding the effectiveness of the PU-prevention
for additional research including development and testing program. Development of a PU in a patient who has been
of a PU risk assessment tool specific to the critical care identified as “at risk” and placed on a prevention protocol
population, either a modified Braden Sale or a newly de- may represent some failure or gap in the PU-prevention
veloped tool. There is precedence for development of a protocol or may represent an unavoidable ulcer.
modified Braden Scale; the Braden Q Risk Assessment Tool The purpose of all PU risk assessment scales is to pre-
is a modified and validated version of the Braden Scale dict PU risk and subsequently mobilize clinicians to imple-
that includes the 6 Braden subscales and an additional ment prevention strategies that will impede PU occurrence.
subscale measuring tissue perfusion and oxygenation; it is Thus, the determination of the predictive validity of a PU
designed for use in the pediatric population.43,44 risk assessment scale cannot be made in isolation from the
Modification of the current Braden Scale may help to prevention strategies that are implemented. However, it is
more accurately identify critical care patients who are at possible that a risk assessment tool designed specifically
significant risk for PU development. Redefining the vari- for the critical care population could help eliminate true
ous levels within the subscale definitions so that they are overprediction and to more accurately identify patients
relevant to the critical care population may be one poten- who need preventive care; thus, further research in this
tial option. The inclusion of other empirically supported area is warranted.
critical care risk factors such as advanced age, prolonged
ICU length of stay, comorbid conditions, or vasopressor
use should undergo further investigation to determine
■ Conclusion
whether their inclusion might enhance the predictive Research indicates that critically ill patients who develop
power of such an instrument in the critical care setting. PUs are classified as at risk by the Braden Scale, and that
Pressure ulcer risk assessment scales, including the most of the critical care patients who did not develop PUs
Braden Scale, tend to overpredict risk; as noted, this may were also classified as at risk. At present, we do not know
be due to an inherent weakness in the tool itself or may whether this discrepancy reflects the success of preventive
reflect the effectiveness of currently used prevention pro- interventions or inadequate discrimination of risk. While
tocols.30 The majority of ICU patients in this review were the subscales Sensory Perception, Mobility, Friction/Shear,
found to be at risk for PU development based on the and Moisture have demonstrated predictive ability in crit-
Braden Scale score but did not develop a PU; it is unknown ical care patients, the paucity of empirical investigations
whether this represents true overprediction or is the result precludes our ability to draw definitive conclusions re-
of preventive care. In the first scenario, overprediction garding the relative contributions of each of these sub-
may be the result of an intrinsic weakness of the scale and scales to PU risk detection and PU development.
results in the unnecessary implementation of prevention Modification of the Braden Scale or development of a crit-
protocols, which could impact healthcare costs. In this ical care PU risk assessment scale might improve our
Copyright © 2012 Wound, Ostomy and Continence Nurses Society™. Unauthorized reproduction of this article is prohibited.
Copyright © 2012 Wound, Ostomy and Continence Nurses Society™. Unauthorized reproduction of this article is prohibited.
31. Bergstrom N, Demuth P, Braden B. A clinical trial of the Braden 38. Benoit R, Watts C. The effect of a pressure ulcer prevention
Scale for predicting pressure sore risk. Nurs Clin North Am. program and the bowel management system in reducing pres-
1987;22(2):417-429. sure ulcer prevalence in an ICU setting. J Wound Ostomy
32. Pancorbo-Hildago T, Garcia-Fernandez F, Lopez-Madina I, Continence Nurs. 2007;34(2):163-175.
Alvarez-Nieto C. Risk assessment scales for pressure ulcer 39. Doughty D. Prevention and early detection of pressure ulcers
prevention: a systematic review. J Adv Nurs. 2006;54(1):94- in hospitalized patients: commentary. J Wound Ostomy
110. Continence Nurs. 2008;35(1):76-78.
33. Seongsook J, Ihnsook J, Younghee L. Validity of pressure ulcer 40. Langer G, Knerr A, Kuss O, Behrens J, Schlomer GJ. Nutritional
risk assessment scales: Cubbin and Jackson, Braden and interventions for preventing and treating pressure ulcers.
Douglas scale. Int J Nurs Stud. 2004;41(2):199-204. Cochrane Database Syst Rev. 2008;(4):CD003216.
34. Cho I, Noh M. Braden Scale: evaluation of clinical usefulness 41. National Pressure Ulcer Advisory Panel. Shear: a contribu-
in an intensive care unit. J Adv Nurs. 2009;66(2):293-302. tory factor in pressure ulcer development. National Pressure
35. Pieper B. Pressure ulcers: impact, etiology and classification. Ulcer Advisory Panel Shear force initiative. http://www.
In:Bryant R, Nix D, eds. Acute and Chronic Wounds: Current npuap.org/Shear_slides.pdf . Published 2005. Accessed
Management Concepts. 4th ed. St Louis, MO: Mosby; 2012:123- March 6, 2011.
135. 42. Compton F, Hoffmann F, Hortig T, et al. Pressure ulcer predic-
36. Defloor T, Grypdonck M. Validation of pressure ulcer risk tors in ICU patients: nursing skin assessment versus objective
assessment scales: a critique. J Adv Nurs. 2004;48(6):613- parameters. J Wound Care. 2008;17(10):417-424.
621. 43. Curley M, Razmus I, Roberts K, Wypij D. Predicting pressure
37. Black JM, Gray M, Bliss DZ, et al. MASD part 2: inconti- ulcer risk in pediatric patients. Nurs Res. 2003;52(1):22-33.
nence-associated dermatitis and intertriginous dermatitis: a 44. Quigley S, Curley M. Skin integrity in the pediatric population:
consensus. J Wound Ostomy Continence Nurs. 2011;38(4):359- preventing and managing pressure ulcers. J Soc Pediatr Nurs.
370. 1996;1(1):7-18.
For more than 70 additional continuing education articles related to skin/wound care and 31
additional continuing education articles related to research, go to NursingCenter.com/CE
Copyright © 2012 Wound, Ostomy and Continence Nurses Society™. Unauthorized reproduction of this article is prohibited.