ADVANCES IN SKIN & WOUND CARE • VOL.18 NO. 2 92 WWW.WOUNDCAREJOURNAL.COM

ORIGINAL INVESTIGATION

Predicting Pressure Ulcer Risk: Comparing thePredictive Validity of 4 Scales

Rostam Jalali, MScN, BSN, and Mansour Rezaie, PhD

comes and increased hospital costs. Health care practitionersexpend much time and energy managing the physical, emo-tional, and financial challenges of treating PrUs. Pressure ulcersimpact patients’ well being, as well as adversely affect morbid-ity and mortality. Facility-acquired PrUs increase a patient’shospital stay, delay a patient’s recuperation, and increase apatient’s risk for developing complications.

Pressure ulcers have been identified as 1 of 7 conditions thataffect a large number of patients and involve relatively expen-sive treatment.2 Yet most PrUs are preventable.3 Because ofthis, strategies are needed for preventing PrUs and containingcosts.2

Reports of PrU incidence vary widely, from 0.4% to 38% inacute care, from 2.2% to 23.9% in long-term care, and from 0%to 17% in home care, according to a report from the NationalPressure Ulcer Advisory Panel4 Prevalence rates show the samevariability: 10% to 18% in acute care, 2.3% to 28% in long-termcare, and 0% to 29% in home care.4 The numbers should beinterpreted cautiously because of discrepancies in methodolo-gy. The most accurate current prevalence rates in the acute caresetting come from 3 multisite studies reported in 2000 and2001: 14.8%, 15%, and 15%.5-7 In the United States, the esti-mated cost associated with approximately 1 to 1.7 millionannual PrUs is $5 to $8.5 billion.8

Prevention of PrUs is thought to be more cost-effective thantreatment.9 Costs for treating a Stage IV PrU can range from$25,000 to $75,000; however, PrU prevention may cost as littleas $500, including nursing time, special mattresses, skin care,and supplemental nutrition.9

The first step in preventing PrUs is determining whichpatients are at high risk for developing them, and for the past30 years, considerable effort has focused on this goal. Since1962, when Doreen Norton designed the Norton Scale,10

researchers and practitioners have continued to search for ascoring system that will predict, with scientific precision, the

Rostam Jalali, MScN, BSN, is a nursing instructor at the College of Nursing, Kermanshah, Iran. Mansour Rezaie, PhD, is a biostatistician at the College of Medicine, Kermanshah, Iran.Acknowledgment: The authors thank the nursing team and patients for their help with and participation in this study. Submitted April 1, 2003; accepted in revised form June 24, 2003.

ABSTRACT

OBJECTIVE: To compare the predictive validity of 4 pressureulcer risk assessment tools.METHODS: Prospective clinical design in which 230 subjectsfree of pressure ulceration on admission were assessed usingthe Braden, Gosnell, Norton, and Waterlow scales within 48hours of admission. Subjects’ skin condition was assessed onceevery 24 hours for a minimum of 14 days to identify any skinbreakdown.RESULTS: Based on Youden’s index, the Gosnell Scale had betterpredictive validity in identifying patients at risk for pressureulcer development (J = 68%). The other scales did not predictindividuals at risk with high accuracy, despite having high sensi-tivity and specificity.CONCLUSION: Numerous pressure ulcer risk assessment toolshave been developed, but sufficient evidence for using one toolover another does not exist. In this study, the Gosnell Scale wasfound to be more appropriate for application in patients withneurologic and orthopedic conditions.CLINICAL RELEVANCE: Being able to predict which patients areat risk for pressure ulcers can assist practitioners in tailoringcare to prevent unnecessary complications and suffering, aswell as reduce costs.

ADV SKIN WOUND CARE 2005;18:92-7.

Pressure ulcers (PrUs) are defined as localized areas of tissuenecrosis that develop when soft tissue is compressed betweena bony prominence and an external surface for a prolongedperiod of time.1 A variety of individuals, including hospital-ized, immobilized, and elderly individuals, are affected byPrUs. Pressure ulcers are associated with negative patient out-

asc1004p000_92-97 03/03/05 1:36 AM Page 92

WWW.WOUNDCAREJOURNAL.COM 93 ADVANCES IN SKIN & WOUND CARE • MARCH 2005

ORIGINAL INVESTIGATION

risk of PrU development. If patients who are at risk for a PrUcan be predicted, their care can be tailored to prevent unnec-essary complications and suffering, as well as save consider-able cost.11

The need for a valid and reliable PrU risk assessment scalethat is applicable to the clinical setting is undeniable. To date,the most commonly adopted scales have been the Braden,12

Gosnell,13 Norton,10 and Waterlow14 scales. Each scale hasbeen reported to have predictive value and applicability in var-ious clinical settings10-14; however, no clinical trial has com-pared the predictive power of these 4 scales.15 The presentstudy sought to do this, and to identify a valid and reliable PrUrisk assessment scale suitable for use in the hospital setting.

METHODSBetween January 2000 and December 2002, a total of 230patients from the neurology, intensive care, orthopedic, andmedical units of 3 educational hospitals in Kermanshah, Iran,participated in the study. Subjects included patients ages 21years and older who had been admitted to the hospital withinthe past 48 hours, who were expected to stay in the hospital for14 days or longer, and who did not have a PrU detected dur-ing the initial skin assessment. Subjects who were expected tostay in the hospital less than 14 days and who had a PrU onadmission were excluded from the study. The 14-day observa-tion period was based on the methods used by Pang andWong,16 whose work suggests that the critical period for PrUdevelopment is the first 14 days after admission.

After consent was obtained from the subjects or their signif-icant others, an initial PrU assessment was performed for eachwithin 48 hours of admission using 4 PrU risk assessmentscales: the Braden, Gosnell, Norton, and Waterlow scales. Afterthe initial assessment, a subject’s skin was assessed once every24 hours for a minimum of 14 days to identify any skin break-down.

InstrumentsThe Braden Scale12 is a summated rated scale comprised of 6subscales: sensory perception, mobility, activity, moisture,nutrition, and friction and shear.The 6 subscales are rated from1 to 4, except the friction and shear subscale, which is ratedfrom 1 to 3. Each numerical rating has a definition of patientcharacteristics to evaluate when assigning a score. A total of 6to 23 points is possible. The original critical cut-off point fordiagnosing risk is 16.12 Other investigators have suggested 18as a cut-off score for older patients and African American andLatino patients.17-19

The Gosnell Scale13 consists of 5 parameters—mental status,continence, mobility, activity, and nutrition—with varying

points (1 to 3 for nutrition; 1 to 4 for continence, mobility, andactivity; and 1 to 5 for mental status). The scoring for eachparameter is clarified by brief descriptive statements. TheGosnell Scale documents additional variables, including bodyvital signs, skin appearance, diet, 24-hour fluid balance, med-ication, and interventions; however, these variables are notgiven weight in the final score. Possible scores for the GosnellScale range from 5 to 20. Although early studies reported thatlower scores denoted a higher risk, 16 has been established asthe critical cut-off score.13

The first PrU risk assessment scale, the Norton Scale10 con-sists of 5 parameters: physical condition, mental state, activi-ty, mobility, and incontinence. Each parameter is rated on ascale from 1 to 4, with a 1-, 2-, or 3-word descriptor for eachrating. The sum of the ratings for all 5 parameters yields ascore ranging from 5 to 20, with lower scores indicating anincreased risk. A score of 16 or lower indicates a risk for PrUformation.10

The Waterlow Scale14 is based on the Norton Scale, but isconsidered to be more comprehensive. The Waterlow Scaleconsists of 8 items: build/weight for height, visual assessmentof the skin in the area at risk, sex and age, continence, mobili-ty, appetite, medication, and special risk factors. The highestand lowest scores of each item vary. For instance, the scores formobility range from 0 to 5; scores for appetite range from 0 to3. Patients scoring 10 to 14 are identified as being at risk forPrU formation. A score of 16 or above is the usual cut-off pointfor at-risk patients in clinical studies.14

Skin assessment chartIn addition to the scales, the research staff used a skin assess-ment tool that delineates bony prominences and requires theassessor to rate the presence or absence of a lesion at each site.A lesion on the skin surface that could be attributed to pres-sure, regardless of the source, was staged according to the fol-lowing criteria20: • Stage I: Nonblanchable erythema of intact skin: the herald-ing lesion of skin ulceration.• Stage II: Partial-thickness skin loss involving the epidermisand possibly the dermis. The ulcer is superficial and presentsclinically as an abrasion, blister, or shallow crater.• Stage III: Full-thickness skin loss involving damage to ornecrosis of subcutaneous tissue that may extend down to, butnot through, underlying fascia. The ulcer presents clinically asa deep crater with or without undermining of adjacent tissue.• Stage IV: Full-thickness skin loss with extensive destruction,tissue necrosis, or damage to muscle, bone, or supportingstructures (ie, tendon or joint capsule). Undermining and sinustracts may also be associated with Stage IV PrUs.

asc1004p000_92-97 03/03/05 1:36 AM Page 93

ADVANCES IN SKIN & WOUND CARE • VOL.18 NO. 2 94 WWW.WOUNDCAREJOURNAL.COM

ORIGINAL INVESTIGATION

A subject’s skin condition was examined according to theskin assessment chart every 24 hours over the period of datacollection.The site and stage of any skin breakdown were indi-cated on a body chart.

A data collection form was used for recording each subject’sgender, age, and treatment characteristics (medications, med-ical problems, and admission date). A nursing interventionchecklist with common preventive and nursing measures wasalso completed with each skin assessment.

ProcedureResearch staff were trained in scoring the PrU risk assessmentscales, staging PrUs, and recording data on the skin assess-ment chart.Videotapes were developed and used in initial andongoing research staff training. Subsequent on-site trainingwas continued until interrater reliability between the investi-gators and research nurses reached an intraclass correlation (r)of 0.95. Interrater reliability between the investigator and theresearch staff was reevaluated at regular intervals and main-tained at r � 0.95 or better throughout the period of data col-lection.

After the assessor gained consent from the subjects or his orher significant other, an initial PrU assessment was performedwithin 48 hours of admission.The 4 PrU risk assessment scaleswere used by 4 independent research nurses for each patient.

Subjects were observed for PrU development every 24 hoursfor a maximum of 14 days. Preventive measures were docu-mented on the nursing intervention checklist.

Data analysisIn the present study, predictive power refers to the ability of ascale to correctly classify patients with characteristics that putthem at high risk for developing a PrU. Predictive power isdetermined by the overall considerations of sensitivity, speci-ficity, positive predictive value, and negative predictive value.

Sensitivity is the percentage of individuals who develop aPrU who were assessed as being at risk for a PrU. A tool hasgood sensitivity if it correctly identifies true positives whileminimizing false negatives.

Specificity is the percentage of individuals who do notdevelop a PrU who were assessed as being not at risk fordeveloping a PrU. A tool has good specificity if it identifies truenegatives while minimizing false positives.

A test has positive predictive value (or a high level of speci-ficity) when the subject’s risk assessment scale score is equal toor less than the cut-off for patients who develop PrUs.Similarly, a test has negative predictive value (or a high level ofsensitivity) when the subject’s risk assessment scale score ishigher than the cut-off for patients who develop PrUs.

The definitions of these measures, which are used in epi-demiology and clinical trials to examine the performance ofscreening or diagnostic tests, are shown in Table 1.

Youden’s index21 (Table 2), proposed in 1950, is an essential-ly equivalent index, in which J � 1�(� � �). If the test has nodiagnostic value, � � 1 � � and J � 0. If the test is invariablycorrect, � � � � 0 and J � 1. Values of J between �1 and 0could arise if the test result was negatively associated with thetrue diagnosis; however, this situation is unlikely to occur inpractice.Youden’s index assumes that sensitivity and specifici-ty have equal importance because � and � are given equalweight. A test with lower values of � and � is more desirablethan a test with higher values.

RESULTSSubjects ranged in age from 21 to 89 years, with a mean age of60 years. One hundred (43.5%) of the subjects were men and130 (56.5%) were women. Seventy-four patients (32.2%) hadPrUs, including 36 women with a mean age of 68.9 years and38 men with a mean age of 68.3 years (Table 3). Pressure ulcerswere found on the sacrum (73%), buttocks (13.5%), heels(8%), and scapula (5.5%); 64.8% of PrUs were Stage II, 24.4%were Stage I, and 10.8% were Stage III. None of the patientshad a Stage IV PrU.

Results of the validity screening tests for the Norton,Waterlow, Gosnell, and Braden scales are summarized inTable 4. Based on the equivalent index assigned in Table 2, theGosnell scale had a better predictive power (Table 5).

A chi-square (�2) test showed a significant differencebetween men and women who developed PrUs (34 women[46%], 40 men [54%]; P � 03). Subjects with a PrU had a meanage of 68.9 years; those without a PrU had a mean age of 64.6

DEFINITIONS OF MEASURES OFPREDICTIVE POWER

Table 1.

Measure Definition

Sensitivity TP

TP + FNx 100

Specificity TN

TN + FPx 100

Positive predictive value TP

TP + FPx 100

Negative predictive value TN

TN + FNx 100

Accuracy TP + TN

Totalx 100

TP = true positive, FN = false negative, TN = true negative, FP = false positive

asc1004p000_92-97 03/03/05 1:36 AM Page 94

WWW.WOUNDCAREJOURNAL.COM 95 ADVANCES IN SKIN & WOUND CARE • MARCH 2005

ORIGINAL INVESTIGATION

years. A 2-tailed independent t test showed a significant dif-ference between the 2 groups (P � .007).

A Pearson product moment correlation test was performedbetween subjects with PrUs and the predictive power of thescales. A significant difference was found between the predic-tive power of the scales (P � .00001). Finally, a 2-tailed inde-pendent t test was conducted between the mean score of sub-jects with or without a PrU and all scales, and a significant dif-ference was found (P � .001).

DISCUSSIONComparison of the predictive validity of 4 commonly used PrUrisk assessment scales showed that the Gosnell Scale had bet-ter predictive validity, based on Youden’s index (68%; Table 5).Sensitivity and specificity for the Gosnell Scale were 85% and83%, respectively. Gosnell and Pontius22 reported sensitivityand specificity of 50% and 73%, respectively. Items included inthe Gosnell scale (mental status, continence, mobility, activity,and nutrition) have been shown to impact PrU formation. Thisscale requires the least amount of time for completion. Becausemost of the subjects in the present study were from neurolog-ic and orthopaedic units, it seems that the Gosnell Scale ismore appropriate in these settings.

In the present study, the Braden Scale had a predictive valueof 53% (Table 5). Its sensitivity, specificity, positive predictivevalue, and negative predictive value were 53%, 100%, 100%,and 58%, respectively. Pang and Wong16 found that the BradenScale had sensitivity, specificity, positive predictive value, andnegative predictive value of 91%, 62%, 100%, and 96%, respec-tively. A study by VandenBosch et al23 showed the specificityand sensitivity of the Braden Scale to be 59% with a cut-off

point of 17. Ramundo24 reported 100% sensitivity and 34%specificity among home care patients with a cut-off score of 18.Capobianco25 reported 71% sensitivity, 83% specificity, 63%positive predictive value, and 88% negative predictive value forthe Braden Scale. A specificity of 100% would be desirable todetermine which patients would not develop a PrU.

The Norton scale had a J equaling 49% in the presentstudy, with 66% accuracy; the positive predictive value andsensitivity were the same as found for the Braden Scale. Chanet al26 reported the Norton Scale to be more appropriate thanthe Waterlow Scale for older patients. Pang and Wong16

found 81% sensitivity, 59% specificity, 33% positive predic-tive value, 93% negative predictive value, and 63% accuracyfor the Norton Scale. This is different from the results in thepresent study, perhaps because Pang and Wong selected sub-jects from rehabilitation units; subjects in the present studywere taken from neurologic, orthopaedic, and critical careunits.

Both Goldstone and Goldstone27 and Lincoln et al28 report-ed a sensitivity of 89% for the Norton Scale. Specificity in the2 studies was vastly different: 36% for Goldstone andGoldstone27 and 94% for Lincoln et al.28 Similar to the BradenScale, the 100% specificity with the Norton Scale is mostappropriate for predicting patients who will not develop aPrU.

In the present study, the Waterlow Scale had a J equaling0.465, with 77% accuracy, 63% sensitivity, 82.5% specificity,

THE YOUDEN’S INDEX

Table 2.

Test

True Test Positive Negative

True positive 1-� = true positive � = false negativeTrue negative � = false positive 1-� = true negative

J = 1-(� + �)

SUBJECTS WITH PRESSURE ULCERS BYSEX AND AGE

Table 3.

With Pressure Ulcer Without Pressure Ulcer

Age Range Men Women Men Women Total

<50 2 4 2 10 1850 to 64 6 4 18 40 6865 to 74 16 16 24 30 8675 to 80 8 12 14 8 42>80 2 4 4 6 16Total 34 40 62 94 230

VALIDITY SCREENING TEST RESULTS

Table 4.

Predictive Power Norton Braden Gosnell Waterlow

Sensitivity (%) 49 53 85 63Specificity (%) 100 100 83 82.5Positive predictive value (%) 100 100 59 61Negative predictive value (%) 52 58 95 84Accuracy (%) 66 71 83 77

DIAGNOSTIC AND SCREENING POWER OF4 SCALES

Table 5.

� � J = 1-(� + �)

Norton .51 0 .49Braden .47 0 .53Gosnell .15 .17 .68Waterlow .37 .16 .465

asc1004p000_92-97 03/03/05 1:36 AM Page 95

ADVANCES IN SKIN & WOUND CARE • VOL.18 NO. 2 96 WWW.WOUNDCAREJOURNAL.COM

ORIGINAL INVESTIGATION

61% positive predictive value, and 84% negative predictivevalue. Compared with the other scales, the Waterlow Scale hadthe lowest scores. These results contrast with reports byHealey29 and Pang and Wong.16 Healey29 reported 100% sen-sitivity and 84% specificity for the Waterlow Scale; Pang andWong16 reported 95% sensitivity and 45% specificity for thisscale. The difference may be accounted for by the diversity ofpatients in the various studies.

A number of factors may be considered limitations of thepresent study, such as nursing care and the support surfacesused by patients. However, these factors were similar for allpatients in the study. The effectiveness of these factors are thesame in the 4 scales and, therefore cannot cause any change inthe results.

Practitioners should use PrU risk assessment scales to helpidentify patients at risk for developing PrUs and institute pre-ventive measures when necessary. However, use of a PrU riskassessment scale alone cannot replace a practitioner’s clinicaljudgment in determining a patient’s risk for PrU develop-ment.30 Pressure ulcers are multidimensional, and other fac-tors, such as mattress type, nursing care, and the patient’snutritional status and other health conditions, may have a rolein PrU development and must be taken into consideration.Patient evaluations must be performed at regular intervals witha PrU risk assessment tool to help prevent PrU formation.11 Inthe present study, PrU prevalence was approximately 32.2%,much higher than PrU prevalence rates reported by otherstudies in acute care. The most current PrU prevalence rates inacute care have been reported as 14.8%, 15%, and 15%.5-7

Subjects in the present study had neurologic problems, such ascerebrovascular accident and intracranial hemorrhage;decreased mobility; decreased level of consciousness; inade-quate nutrition; and incontinence, which may account for thehigher PrU prevalence rate. In addition, nursing care in Iran isnot up-to-date, and air mattresses and other pressure-relievingor pressure-reducing equipment are not used. Again, that couldexplain the higher prevalence rate found in the present study.

The age range of patients who developed PrUs in the pres-ent study is comparable to other studies. Hausman9 reportedthat 90% of patients who developed a PrU were older than 65years. Because life expectancy in Iran is lower than in westerncountries, the mean age of 68 years found in this study couldbe considered higher than that found in western countries.

Pressure ulcer prevalence in the first and second week ofhospitalization was found to be the same as reported by Pangand Wong,16 with 34% of PrUs occurring in the first week and70% occurring in the second week.

Schue and Langemo31 reported PrU prevalence by stage tobe 57% for Stage II, 24% for Stage I, and 15% for Stage III. For

patients in a supine position, the most common sites of PrUswere the sacrum, buttocks, heel, and occiput, which is compa-rable to the findings of the present study. According to Leeet al32, 96% of PrUs occur on the lower limbs, including 67%on the sacrum and buttocks and 29% on the heels. Leigh andBennett33 found the sacrum to be the most common site forPrUs (60%). Schue and Langemo32 had a similar finding, with46% of PrUs occurring in this area.

Many interventions can help prevent PrUs, including the useof electrical air mattresses, pedal boards, sheepskin, pillows,limb massage, and creams and changing a patient’s sheets andbody position. In the present study, subjects received routinenursing care in combination with a turning regimen to com-plement the range of multidisciplinary activities needed tomanage patients at risk for PrUs. Because the patient care pro-tocol in the present study was consistent, these interventionscould not have influenced the results of the study.

CONCLUSIONNumerous PrU risk assessment tools have been developed andused in clinical practice. However, sufficient evidence for usingone tool over another does not exist. The present study soughtto compare the predictive validity of 4 PrU risk assessmenttools: the Braden Scale, the Gosnell Scale, the Norton Scale,and the Waterlow Scale. The Gosnell Scale was found to bemore appropriate for application in 2 of the primary patientgroups evaluated in this study, those with neurologic condi-tions and those with orthopedic conditions. ●

REFERENCES1. Reid J, Morison M. Classification of pressure sore severity. Nurs Times 1994;90:46-50.2. Layon AJ, George BE, Hamby B, Gallagher TJ. Do elderly patients overutilize health care

resources and benefit less from them than younger patients? A study of patients whounderwent craniotomy for treatment of neoplasm. Crit Care Med 1995;23:829-34.

3. National Pressure Ulcer Advisory Panel. Statement on pressure ulcer prevention, 1992.Available online at http://www.npuap.org; accessed March 18, 2004.

4. National Pressure Ulcer Advisory Panel Board of Director. Pressure ulcers in America: preva-lence, incidence, and implications for the future. Adv Skin Wound Care 2001;14:208-15.

5. Amlung S, Muller W, Bosley LM. The 1999 National Pressure Ulcer Prevalence Survey: abenchmarking approach. Adv Skin Wound Care 2001;14:297-301.

6. CalNOC. A statewide nursing outcomes database. Linking patient outcomes to hospitalnursing care: California Nursing Outcomes Coalition; January 8, 2000.

7. Whittington K, Patrick M, Roberts JL. A national study of pressure ulcer prevalenceand incidence in acute care hospitals. J Wound Ostomy Continence Nurs 2000;27:209-15.

8. Beckrich K, Aronovitch SA. Hospital-acquired pressure ulcers: a comparison of costs inmedical vs. surgical patients. Nurs Econ 1999;17:263-71.

9. Hausman LL. Cost containment through reducing pressure ulcers. Nurs Manage 1994;25:88R, 88T, 88V.

10. Norton D, et al. An investigation of Geriatric Nursing Problems in Hospital. London: NationalCorporation for the Care of Old People; 1962.

11. Waterlow J. Pressure sore assessments. Nurs Times 1996;92:53-6, 58.12. Bergstrom N, Braden BJ, Laguzza A, Holman V. The Braden Scale for Predicting Pressure

Sore Risk. Nurs Res 1987;36:205-10.13. Gosnell DJ. An assessment tool to identify pressure sores. Nurs Research 1973;22(1):55-9.

asc1004p000_92-97 03/03/05 1:36 AM Page 96

WWW.WOUNDCAREJOURNAL.COM 97 ADVANCES IN SKIN & WOUND CARE • MARCH 2005

ORIGINAL INVESTIGATION

14. Waterlow J. Pressure sores: a risk assessment card. Nurs Times 1985;81(48):49-55.15. Edwards M. Pressure sore risk calculators: some methodological issues. J Clin Nurs

1996;5:307-1216. Pang S, Wong TK. Predicting pressure sore risk with the Norton, Braden, and Waterlow

scales in a Hong Kong rehabilitation hospital. Nurs Res 1998;47:147-53.17. Bergstrom N, Braden B. A prospective study of pressure sore risk among institutionalized

elderly. J Am Geriatr Soc 1992;40:747-5818. Lyder CH, Yu C, Stevenson D, et al. Validating the Braden Scale for the prediction of pres-

sure ulcer risk in blacks and Latino/Hispanic elders: a pilot study. Ostomy Wound Manage1998;44(suppl 3A):42S-49S.

19. Lyder CH, Yu C, Emerling J, et al. The Braden Scale for pressure ulcer risk: evaluating andpredicting validity in Black and Latino/Hispanic elders. Appl Nurs Res 1999;12(2):60-8.

20. Bergstrom N, et al. Treatment of Pressure Ulcers in Adults. Clinical Practice Guideline No.15. Rockville, MD: Agency for Health Care Policy and Research; 1994.

21. Armitage L. Statistical Methods in Medical Research. 4th ed. New York: BlackwellScientific; 2001. p 692-8.

22. Gosnell DJ, Pontius C. A model of quality assurance for decubitus ulcer monitoring.Decubitus 1988;1(4):24-9.

23. VandenBosch T, Montoye C, Satwicz M, Durkee-Leonard K, Boylan-Lewis B. Predictivevalidity of the Braden Scale and nurse perception in identifying pressure ulcer risk. ApplNurs Res 1996;9:80-6.

24. Ramundo JM. Reliability and validity of the Braden Scale in the home care setting. J WoundOstomy Continence Nurs 1995;22:128-34.

25. Capobianco ML, McDonald DD. Factors affecting the predictive validity of the BradenScale. Adv Wound Care 1996;9(6):32-6.

26. Chan WH, Chow KW, French P, Lai YS, Tse LK. Which pressure sore risk calculator? Astudy of the effectiveness of the Norton scale in Hong Kong. Int J Nurs Stud 1997;34:165-9.

27. Goldstone LA, Goldstone J. The Norton score: an early warning of pressure sores? J AdvNurs 1982;7:419-26.

28. Lincoln R, et al. Use of Norton Pressure Sore Risk Assessment Scoring System with elder-ly patients in acute care. J Enterostomal Ther 1986;13:132-8.

29. Healey F. Waterlow revisited. Nurs Times 1996;92:80-4.30. Bergstrom N, Demuth PJ, Braden BJ. A clinical trial of the Braden Scale for Predicting

Pressure Sore Risk. Nurs Clin North Am 1987;22:417-28.31. Schue RM, Langemo DK. Pressure ulcer prevalence and incidence and a modification of

the Braden Scale for a rehabilitation unit. J Wound Ostomy Continence Nurs 1998;25:36-43.

32. Lee BY, Burton L, Herz L. Surgical Management of Cutaneous Ulcers and Pressure Sores.New York: Chapman and Hall; 1998:1-4, 86-116.

33. Leigh IH, Bennett G. Pressure ulcers: prevalence, etiology, and treatment modalities. Areview. Am J Surg 1994;167:25S-30S.

asc1004p000_92-97 03/03/05 1:36 AM Page 97