I N T EG R A T I V E L I T E R A T U R E R E V IE W S A N D M E T A - A N A L Y S E S

Risk assessment scales for pressure ulcer prevention: a systematic review Pedro L. Pancorbo-Hidalgo

PhD RN

Professor of Medical-Surgical Nursing, School of Nursing, University of Jae´n, Jae´n, Spain

Francisco Pedro Garcia-Fernandez

RN

Manager of Training, Research and Quality Unit, University Hospital of Jae´n, Jae´n, Spain

Isabel Ma Lopez-Medina

PhD Student RN

Professor of Medical-Surgical Nursing, School of Nursing, University of Jae´n, Jae´n, Spain

Carmen Alvarez-Nieto

PhD RN

Professor of Mother and Childhood Nursing, School of Nursing, University of Jae´n, Jae´n, Spain

Accepted for publication 17 January 2005

Correspondence: Pedro L. Pancorbo-Hidalgo, School of Nursing, University of Jae´n, Las Lagunillas s/n 23071, Jae´n, Spain. E-mail: [email protected]

94

PANCORBO-HIDALGO P.L., GARCIA-FERNANDEZ F.P., LOPEZ-MEDINA I.M. & A L V A R E Z - N I E T O C . ( 2 0 0 6 ) Journal of Advanced Nursing 54(1), 94–110 Risk assessment scales for pressure ulcer prevention: a systematic review Aim. This paper reports a systematic review conducted to determine the effectiveness of the use of risk assessment scales for pressure ulcer prevention in clinical practice, degree of validation of risk assessment scales, and effectiveness of risk assessment scales as indicators of risk of developing a pressure ulcer. Background. Pressure ulcers are an important health problem. The best strategy to avoid them is prevention. There are several risk assessment scales for pressure ulcer prevention which complement nurses’ clinical judgement. However, some of these have not undergone proper validation. Method. A systematic bibliographical review was conducted, based on a search of 14 databases in four languages using the keywords pressure ulcer or pressure sore or decubitus ulcer and risk assessment. Reports of clinical trials or prospective studies of validation were included in the review. Findings. Thirty-three studies were included in the review, three on clinical effectiveness and the rest on scale validation. There is no decrease in pressure ulcer incidence was found which might be attributed to use of an assessment scale. However, the use of scales increases the intensity and effectiveness of prevention interventions. The Braden Scale shows optimal validation and the best sensitivity/ specificity balance (57Æ1%/67Æ5%, respectively); its score is a good pressure ulcer risk predictor (odds ratio ¼ 4Æ08, CI 95% ¼ 2Æ56–6Æ48). The Norton Scale has reasonable scores for sensitivity (46Æ8%), specificity (61Æ8%) and risk prediction (OR ¼ 2Æ16, CI 95% ¼ 1Æ03–4Æ54). The Waterlow Scale offers a high sensitivity score (82Æ4%), but low specificity (27Æ4%); with a good risk prediction score (OR ¼ 2Æ05, CI 95% ¼ 1Æ11–3Æ76). Nurses’ clinical judgement (only considered in three studies) gives moderate scores for sensitivity (50Æ6%) and specificity (60Æ1%), but is not a good pressure ulcer risk predictor (OR ¼ 1Æ69, CI 95% ¼ 0Æ76–3Æ75). Conclusion. There is no evidence that the use of risk assessment scales decreases pressure ulcer incidence. The Braden Scale offers the best balance between sensitivity and specificity and the best risk estimate. Both the Braden and Norton


2006 Blackwell Publishing Ltd, Journal of Advanced Nursing

Integrative literature reviews and meta-analyses

Risk assessment scales for pressure ulcer prevention

Scales are more accurate than nurses’ clinical judgement in predicting pressure ulcer risk. Keywords: meta-analysis, nursing, pressure ulcers, risk assessment scales, risk factors, systematic review

Introduction Pressure ulcers (PUs) constitute a health problem of utmost importance, given the number of people affected worldwide (Lyder 2002). A recent study on PU epidemiology in Spain has recorded the prevalence of this type of lesion: 8Æ24% of hospitalized patients, 8Æ34% of patients with home care services and 6Æ43% of patients in long-stay health centres (Torra i Bou et al. 2003). The situation is similar in other developed countries, with prevalence rates ranging from 3% to 30%, and an incidence of 1% to 50% (Bergstrom et al. 1994, Smith 1995, Colin et al. 1997, Soldevilla Agreda & Torra i Bou 1999; Thoroddsen 1999, Maklebust & Sieggreen 2001, Lyder 2002, Torra i Bou et al. 2003). The mortality rate directly attributed to PU is quite high – 5Æ283 deaths between 1Æ987 and 1Æ999 in Spain (Verdu´ et al. 2003) – which means that almost 400 people die every year because of PUs. This problem also affects healthcare services, since PUs generate substantial financial costs. Some studies estimate that the cost of treatment per patient ranges between $1200 and 12,000 (Colin et al. 1997) and between $500 and 100,000 (Lyder 2003), depending on the location studied, and an annual total cost of between $1Æ300 million (Maklebust & Sieggreen 2001) to $3Æ100 million (Lyder 2003). In Spain, the annual cost of PU treatment is estimated to be around $1Æ687 million, which is 5Æ20% of the total healthcare expenditure, and a little higher than in the United Kingdom (UK), where it is estimated to be 3Æ20% of the total healthcare expenditure (Lyder 2003). Prevention is usually considered the most efficient method to tackle the problem [Bergstrom et al. 1992, Land 1995, European Pressure Ulcer Advisory Panel (EPUAP) 1998, Pancorbo Hidalgo & Garcı´a Ferna´ndez 2002, GNEAUPP 2000], and the quality of nursing care is considered the key factor in dealing with PUs. Prevention measures are generally divided into four major areas: assessment of PU development risk, skin care and initial treatment, use of pressure-reducing support surfaces, and education (Bergstrom et al. 1992, Garcia Ferna´ndez et al. 2001). The purpose of risk assessment is to identify those requiring prevention measures and their specific risk factors. One recommendations for clinical practice is to perform a systematic risk assessment by means of a validated scale, such as the Braden Scale (Bergstrom et al.
2006 Blackwell Publishing Ltd, Journal of Advanced Nursing

1992, Best Practice 1997; Research Dissemination Core 2002, The Nursing Best Practice Guideline 2003) or the Norton Scale (Bergstrom et al. 1992, Best Practice 1997). A risk assessment scale (RAS) for PU prevention is a tool for establishing a score according to a series of parameters considered to be risk factors. Nowadays, several RASs are used in clinical practice, although most have not yet been properly validated (Thomas 2001, Torra i Bou 1997). Some of the most widely used RASs are the Norton Scale and its modifications, the Braden Scale and, in the UK, the Waterlow Scale. Nevertheless, there is no consensus on each scale’s cutoff scores, so that there are no generally accepted indications whether at-risk patients will develop PUs and whether prevention measures should be adopted. In other words, these scales do not prevent the application of prevention measures to patients not in need of them or not applying them to vulnerable patients (Smith et al. 1995, Thomas 2001). In addition, the use of these assessment scales in clinical practice is still limited and is frequently replaced by the nurses’ subjective assessment or clinical judgement, even though this method has not been validated (VandeBosch et al. 1996).

Objectives The objectives of this research were • to determine the effectiveness of the use of RASs in clinical practice for PU prevention, measured in terms of (a) decrease in PU incidence, (b) more efficient use of PU prevention interventions; • to determine the degree of validation of different RASs in the bibliography; • to determine the effectiveness of various RASs as indicators of a patient’s risk of developing PU.

Search methods Search strategy We conducted a bibliographic search using 14 bibliographic databases: • Database of Abstracts of Reviews of Effectiveness (DARE); • CINAHL; • Medline; 95

P.L. Pancorbo-Hidalgo et al.

• Current contents: Clinical Medicine, Social and Behavioural Sciences, Life Sciences; • Indice me´dico espan˜ol (IME, Spanish Medical Index); • Cuiden; • Centro Latinoamericano y del Caribe de Informacio´n en Ciencias de la Salud (LILACS, Latin-American and Caribbean Information Centre for Health Sciences); • Cochrane Library; • EBSCO; • ScienceDirect; • Springer; • InterSciencia; • ProQuest; • Pascal. The period included in the search was 1966–2003, and we used the following terms as keywords: pressure ulcers or pressure sores or decubitus ulcer and risk assessment; in some of the databases we combined them with the term clinical trial. We selected documents written in Spanish, English, French and Portuguese. Besides, and also searched the bibliographic references of the selected studies. In order to locate and retrieve unpublished studies, we studied research reports, proceedings from national and international conferences, and we contacted experts in the Spanish Group for PUs Study and Assessment.

Inclusion criteria The inclusion criteria for the review were • Studies of the use of RAS for PU prevention offering results on PU incidence. • Design of research: controlled clinical trials and prospective cohort studies. Retrospective studies were not included, given their greater likelihood of bias and unreliability. • In prospective cohort studies, the patients considered had not developed PU at the beginning of the study; this guarantees the scale’s predictive validity. • The percentage of patients followed up during the specified period was over 75%. In other words, the drop-out rate of patients did not exceed 25% in accordance with established criterion (Sackett et al. 1997). Patients were followed up in a systematic way during an established period. • Studies offering data on the predictive validity of the scales (sensitivity and specificity) or the raw data which are necessary to do the calculations. Sensitivity is the percentage of subjects who develop PUs and that the scale predicts would develop them. Specificity is the percentage of those who do not develop PUs and that the scale does not predict would develop them. 96

• Studies published in any of the following languages: Spanish, English, French and Portuguese.

Exclusion criteria Studies were excluded in they used the same data to generate the scale and to establish its validity, since sensitivity and specificity values are unreliable in such cases.

Assessment of methodological quality Critical assessment of the studies retrieved was based on the following: • For clinical trials, we used the CASP Guide (Critical Appraisal Skills Programme 2002) For prospective cohort studies, we used the critical assessment guide developed for the clinical practice guide for PU assessment and prevention (Rycroft-Malone & McInness 2000). Two independent researchers reviewed each study selected. In the event of disagreement over its validity, a third researcher reviewed the study, and decided whether or not it should be accepted as valid (Ga´lvez Toro 2001).

Data extraction The data from each selected study were transferred to a data extraction sheet. Two researchers independently extracted the data from each valid study. In order to minimize bias, a list was drawn up of operational definitions of the variables considered. Whenever possible, we re-calculated the validity indicator scores (sensitivity, specificity, positive prediction value, negative prediction value, effectiveness and area under the ROC curve) in order to check them against the raw data offered by each author. If one of these indicators, or the effectiveness indicator, was missing in the original study, we used the score calculated by the reviewers on the basis of the original data. We calculated the odds ratio (OR) from the data offered by the original studies.

Indicators analysis For RASs having two or more valid original studies, we conducted an overall study of the indicators considered. For each RAS, we calculated the weighted average values of the validity indicators, using the inverse of variance as the method for weighting. Studies with a higher variance or smaller samples were given less weight. We also performed a meta-analysis of the magnitude effect indicators (OR) by using the random effects model (DerSimonian-Laird
2006 Blackwell Publishing Ltd, Journal of Advanced Nursing

Integrative literature reviews and meta-analyses

model) with the statistic package SPSS, and the macroMAR v2001 by Bonillo, Dome´nech & Granero (Delgado 2002).

Findings We found 491 bibliographical references in our search, out of which we selected 79 after a critical reading of the abstracts. We then selected 33 as valid, using the above criteria. The remaining 46 studies were considered non-valid for various reasons: they did not provide information to validate the scales used; the design of the studies did not meet our criteria; they had an excessive number of patient drop-outs during the follow-up period.

Clinical effectiveness We found three experimental studies assessing the clinical effectiveness of RASs for PU prevention, all using only the Norton Scale and its modifications. Table 1 shows the design of these studies and the assessment scales used. Table 2 shows their main results.

Validity Table 3 shows the main features of the validation studies on the use of RASs for PU prevention. The most widely studied scale was the Braden Scale, which was analysed in 22 studies, followed by the Waterlow Scale, which was used in seven studies. The following scales were studied in fewer articles, in descending order: Norton, Cubbin–Jackson, Modified Norton Scale (MNS), Andersen, Douglas, Knoll, Pressure Sore Prediction Score (PSPS), Risk Assessment Pressure Sore (RAPS), Fragmment and Emina. Only three studies used both nurses’ clinical judgement and the Braden Scale (Salvadalena et al. 1992, VandeBosch et al. 1996) or the Waterlow or Norton Scales (Smith 1989). The centres where the studies were conducted were mainly hospital units, and the scales were validated less often in other medical assistance contexts, such as home care services (10Æ34%) – where the Braden and the Waterlow Scales were used – and elder care centres (6Æ89%) – where the Norton, Waterlow and Knoll Scales were used. The studies tended to use systematic sampling of consecutive admissions (44Æ82%) more often than convenience (34Æ48%) and random sampling (20Æ68%). Sample size were very heterogeneous, so that weighting of the figures according to the inverse of the variance proved necessary in order to compare scale validity in the different studies. The number of patient drop-outs at the end of study periods was not
2006 Blackwell Publishing Ltd, Journal of Advanced Nursing

Risk assessment scales for pressure ulcer prevention

excessive, since that was one of the exclusion factors for the initial selection of the papers. The follow-up period established by the studies had a wide time span, ranging from 5 days in Lewicki et al. (2000) to 12 weeks (Bergstrom et al. 1987b, Goodridge et al. 1998, Schoonhoven et al. 2002, Lindgren et al. 2002). Some other factors were considered, including discharges from hospital, deaths or transfers, and complete period of stay in the unit. The average age of patients in the studies analysed varied between 50Æ5 years in Bergstrom et al. (1987a) and 83Æ1 years in Edwards (1995). One of the inclusion criteria in most studies was that patients should be over a certain age, around 60 years. Some studies did not give information about patients’ age. The highest PU incidence found was 46Æ6% in patients in an elder care centre, whose average age was 81Æ3 years (Towey & Erland 1988). At the other extreme was an incidence of 1Æ17% in an emergency unit (Andersen et al. 1982). This low incidence in a sizeable sample (n ¼ 3398) may be accounted for by the characteristics of the unit and the fact the follow-up period was only 10 days. Regarding stage of development of the PUs considered in the studies, stage I was considered to be the minimum, except by Ramundo (1995), Schoonhoven et al. (2002) and Hagisawa and Barbenel’s (1999), who considered stage II as the minimum. With respect to masking risk data from the researchers who were to determine the patients’ skin condition, this was carried out in only eight studies (Table 3), whereas masking from the nurses in charge of patients was performed in another eight studies. Double masking was performed in only four studies. In Gunningberg et al. (1999), only masking of nurses in the control group was performed. Different studies gave different cut-off scores for risk of developing a PU, although the scale and healthcare contexts were the same. For the Braden Scale in acute hospitals, some authors used a rate £16 to establish PU risk, while others set this at below 18. In intensive care units, the most widely used rate was £16, whereas the cut-off score was £18 for long-stay patients in elder care centres, long-term patients or those having home care services. For the Norton Scale, there was less variability and the most frequent cut-off score was £16. The most usual cut-off score for the Waterlow Scale was that proposed by the author, that is, ‡10, although some studies used a remarkably higher rate in an attempt to improve specificity.

Braden Scale We found 22 studies with validation data on the Braden Scale (Table 4), and therefore we can consider it as having optimal validation. The inter-rater reliability is high 97

98

UK

Bale et al. (1995)

Before and after: stage I (control); intervention; stage 2 (experimental)


Before and after: stage I (control); stage 2 (experimental)

Controlled clinical trial, without random assignment. Partial masking (control group) from the nurses

Design (a) To investigate PU prevalence in patients with hip fracture; (b) to check if the clinical use of the Modified Norton Scale can identify at-risk patients of PU; (c) to compare PU prevalence in an experimental group as opposed to a control group To check whether the standardized and recurrent assessment of PU risk, together with the use of the adequate pressure relief system decreases the PU incidence* To assess the effect of the use of the Norton Scale in the care plan on clinical practice and the patients’ skin condition

Objective of the study

Norton

Modified Norton Scale (six factors) Cut-off scores £10 A total of 223 patients Experimental group: 104 (79 patients without PU when admitted) Control group: 223 (161 patients without PU when admitted) A total of 181 patients Experimental group: 89; control group: 92 Non-random. Consecutive admissions of patients between May 1991 and December 1993

Palliative care centres

Non-random. Patients admitted in one of the four units, with a stay longer than 10 days, over 18 years and willing to participate

Modified-Norton Scale Cut-off score: 3 Braden £16 Cubbin–Jackson £24 Douglas £ 18

Perneger et al. (2002)

Braden £16 Norton £16 Waterlow ‡10

Lindgren et al. (2002)

Hospital (medical, surgical and geriatric units) Hospital (medical, surgical, orthopaedic and geriatric units) Hospital (medical surgical units, ICU) Hospital (ICU)

673

Random

Convenience

Systematic

Convenience

125

1190

530

Convenience 1431

314

1711

Systematic

Systematic

337

320

Systematic

Systematic

84

Systematic

Sampling

13

0

42

202

0

0

15

0

0

10

62

61Æ4

69Æ2

60Æ1

ND

57Æ5

78Æ8

62

60Æ9

72

I

I

II

I

I

I

I

I

I

Not

Not

Not

Yes



Yes§

Not

Yes



ND

ND

Not

Yes
ND

Not

ND

Not

Yes

Yes§

Yes**

Yes

ND

ND

ND

Not

Yes

ND

Not

Not

ND

Not

ND

Minimum PU Blinding to Blinding to stage prevention researchers caring nurses

To discharge/ I moved to other ward/death

3 weeks/to discharge

12 weeks

12 weeks/to discharge

7 days

ICU stay

8 weeks

5 days

ND

To discharge

Average age Follow-up Sample Lost to size follow-up (years) period

ND, no data; ICU, intensive care unit. *Consecutive admissions.
Standard care to all patients.  Prevention measures in patients with stage I PU. § Pressure-relief mattress and regular repositioning to all patients. – Regular repositioning. Patient with pressure-relief mattress are not included in the study. **Preventive measures for risk patients.

Standard care. Data from patients with and without prevention are separately analysed.

Seongsook et al. (2004)

Hospital (medical surgical units) Hospital (medical surgical units) Hospital (cardiac surgery) Home care

Setting

Waterlow ‡10 Hospital (ICU) Cubbin–Jackson £29 Emina ‡4 Hospital

Braden £19

Braden£16

Lyder et al. (1999)

Authors

RAS and cut-off score

Table 3 (Continued)

Integrative literature reviews and meta-analyses Risk assessment scales for pressure ulcer prevention

103

104

Braden Scale Bergstrom et al. (1987a) Bergstrom et al. (1987a) Bergstrom et al. (1987b) Langemo et al. (1991) (hospital) Langemo et al. (1991) (long-term care) Salvadalena et al. (1992) Barnes and Payton (1993) Braden and Bergstrom (1994) Ramundo (1995) Capobianco and McDonald (1996) VandeBosch et al. (1996) Pang and Wong (1998) Goodridge et al. (1998) Bergstrom et al. (1998) Lyder et al. (1999) Hagisawa and Barbenel (1999) Halfens et al. (2000) Lewicki et al. (2000) Bergquist and Frantz (2001) Schoonhoven et al. (2002) Perneger et al. (2002) Seongsook et al. (2004) Norton Scale Stotts (1988) Smith (1989) Wai-Han et al. (1997) Pang and Wong (1998) Schoonhoven et al. (2002) Perneger et al. (2002) Waterlow Scale Smith (1989) Edwards (1995) Wai-Han et al. (1997) Pang and Wong (1998) Westrate et al. (1998) Boyle and Green (2001) Schoonhoven et al. (2002)

Author

30 9Æ7 4Æ32 20 7Æ9 5Æ2 11

17 30 4Æ32 20 11 15

100%§ ND ND 0Æ99 ND ND

£14 £16 £14 £16 £16 ND ND 92Æ5%§ ND 0Æ99 ND ND ND

20 6 27Æ5 15 28 28Æ1 20 9Æ7 12Æ8 32Æ4 4Æ4 14Æ7 4Æ7 6Æ3 11 15 31Æ3

0Æ90 0Æ86 0Æ99 0Æ83 ND 90%§ 0Æ99 ND ‡0Æ95 0Æ99 ND 0Æ86 0Æ90 ND ND ND ND

£18 £16 £18 £18 £18 £17 £18 £19 £18 £16 £16 £20 £14 £19 £16 ND £16

‡10 ND 10 ‡16 ‡15 ‡10 ‡10

9 7 40 14 28

0Æ99 0Æ99 0Æ89 ND ND

Reliability*

Pressure ulcer incidence (%)

£16 £16 £16 £16 £18

Cut-off score

75Æ8 100 87Æ5 95 80Æ9 100 89Æ5

16 62 75 81 46Æ2 ND

60Æ0 72Æ7 78Æ6 100Æ0 71Æ4 58Æ6 90Æ5 50Æ0 51Æ9 77 38Æ9 61Æ7 66Æ6 61Æ0 43Æ5 ND 97Æ0

100Æ0 100Æ0 83Æ3 54Æ5 57Æ1

Sensitivity (%)

Table 4 Reliability, validity and risk prediction of pressure ulcer risk assessment scales

38 10Æ3 28Æ8 44 28Æ5 13 22Æ4

94 31 66Æ6 59 60Æ4 ND

54Æ4 90Æ6 74Æ3 34Æ1 83Æ3 40Æ5 62Æ4 52Æ3 77Æ8 50 100Æ0 79Æ9 29Æ6 68Æ0 67Æ8 ND 26Æ0

90Æ2 63Æ7 63Æ9 93Æ7 61Æ1

Specificity (%)

33Æ3 7Æ1 5Æ3 29 8Æ9 ND 6Æ7

38 26Æ9 9Æ23 33 7Æ1 ND

25Æ0 33Æ3 53Æ7 20Æ6 62Æ5 27Æ9 37Æ3 10Æ1 25Æ6 77 100Æ0 34Æ5 4Æ5 11Æ4 14Æ3 ND 37Æ3

43Æ8 21Æ4 60Æ6 60Æ0 36Æ4

Positive predictive value (%)

65Æ7 100 98Æ1 97 94Æ5 ND 97Æ2

84 64Æ7 98Æ3 93 94Æ5 ND

84Æ3 98Æ1 90Æ2 100Æ0 88Æ2 71Æ4 96Æ4 90Æ7 91Æ7 50 91Æ6 92Æ4 94Æ7 96Æ3 90Æ7 ND 95Æ0

100Æ0 100Æ0 85Æ2 92Æ2 78Æ6

Negative predictive value (%)

44Æ6 16Æ1 31Æ4 54 32Æ6 ND 29Æ8

80Æ5 39Æ6 68Æ1 63 58Æ8 ND

55Æ6 89Æ5 75Æ5 43Æ8 80Æ0 45Æ6 67Æ9 52Æ1 74Æ5 ND 92Æ0 77Æ2 31Æ4 67Æ6 65Æ1 ND 48Æ2

90Æ9 67Æ0 71Æ7 87Æ8 60Æ0

Efficacy (%)


ND ND ND ND ND 0Æ66 0Æ61

ND ND ND ND 0Æ56 0Æ74

ND ND ND ND ND ND ND ND ND ND ND ND ND ND 0Æ55 0Æ74 0Æ71

ND ND ND ND ND

ROC

0Æ96 ND 2Æ83 15Æ42 1Æ68 ND 2Æ46

2Æ98 0Æ67 6Æ32 6Æ07 1Æ31 ND

1Æ76 25Æ58 10Æ61 ND 12Æ5 0Æ97 15Æ73 1Æ10 3Æ78 ND ND 6Æ39 0Æ84 3Æ32 1Æ62 ND 11Æ36

(1Æ39–4Æ34)

(0Æ34–23Æ61) (1Æ98–120Æ19) (0Æ80–3Æ57)

(0Æ40–2Æ28)

(1Æ34–6Æ64) (0Æ28–1Æ63) (1Æ24–32Æ27) (1Æ88–19Æ60) (0Æ91–1Æ88)

(1Æ53–84Æ6)

(3Æ31–12Æ33) (0Æ29–2Æ43) (2Æ22–4Æ97) (1Æ13–2Æ33)

(2Æ92–53Æ48) (0Æ40–2Æ31) (3Æ44–72Æ06) (0Æ53–2Æ28) (2Æ49–5Æ73)

(0Æ66–4Æ86) (9Æ53–69Æ99) (3Æ74–30Æ11)

ND ND 8Æ85 (2Æ48–31Æ53) 17Æ70 (3Æ72–84Æ24) 2Æ10 (0Æ36–12Æ32)

Odds ratio (CI 95%)

P.L. Pancorbo-Hidalgo et al.


2006 Blackwell Publishing Ltd, Journal of Advanced Nursing


2006 Blackwell Publishing Ltd, Journal of Advanced Nursing 29Æ6

ND

£21

31Æ3

ND ND ND ND

£18 Yes/No Yes/No Yes/No

30 20 28Æ1

15

11Æ7

5Æ2 31Æ3

ND

0Æ83

ND ND

>3

£36

£29 £24

7

46Æ6

ND

>12

0Æ93

4Æ2

ND

>6

‡4

1Æ17

ND

Reliability*

Pressure ulcer incidence (%)

‡2

Cut-off score

ND, no data. *Correlation coefficient, Pearson’ r.
Percentage of correctly classified patients.  Area under Receiver–Operator Curve. § Percentage of observer agreement. – PSPS, Pressure Sore Prediction Score. **RAPS, Risk Assessment Pressure Sore.

Andersen Scale Andersen et al. (1982) PSPS– Lowthian (1989) Knoll Scale Towey and Erland (1988) Modified Norton Scale Gunningberg et al. (1999) Emina Scale Fuentelsaz Gallego (2001) Cubbin–Jackson Scale Boyle and Green (2001) Seongsook et al. (2004) RAPS** Lindgren et al. (2002) Fragmment Scale Perneger et al. (2002) Douglas Scale Seongsook et al. (2004) Clinical judgement Smith (1989) Salvadalena et al. (1992) VandeBosch et al. (1996)

Author

Table 4 (Continued)

50Æ0 50 51Æ7

100

62

57Æ4

83 89

76Æ6

71

85Æ7

89

88

Sensitivity (%)

47Æ9 79Æ7 58Æ9

18

85

57Æ6

42 61

71Æ6

44

56Æ3

76

87

Specificity (%)

28Æ8 38Æ5 33Æ3

34

33Æ5

14Æ4

ND 51

16Æ9

35

63Æ2

14Æ2

7

Positive predictive value (%)

69Æ4 86Æ3 75Æ4

100

94Æ9

91Æ6

ND 92

97Æ6

78

81Æ8

99Æ4

100

Negative predictive value (%)

48Æ5 73Æ7 56Æ9

43Æ6

ND

57Æ4

ND 69Æ8

71Æ9

52Æ0

70Æ0

76Æ6

83Æ8

Efficacy (%)


ND ND ND

0Æ79

0Æ79

ND

0Æ72 0Æ82

0Æ82

ND

ND

ND

ND

ROC

0Æ92 (0Æ39–2Æ16) 3Æ94 (1Æ40–11Æ07) 1Æ54 (0Æ65–3Æ65)

ND

ND

1Æ82 (1Æ06–3Æ11)

ND 12Æ66 (3Æ99–40Æ11)

8Æ24 (4Æ10–16Æ54)

1Æ92 (0Æ69–5Æ36)

7Æ71 (2Æ17–27Æ42)

25Æ62 (10Æ73–61Æ20)

36Æ07 (14Æ07–92Æ45)

Odds ratio (CI 95%)

Integrative literature reviews and meta-analyses Risk assessment scales for pressure ulcer prevention

105

P.L. Pancorbo-Hidalgo et al.

(MNS, Emina), others from a combination of factors considered in other scales (RAPS, Fragmment), while yet others were developed in an original way (Andersen, PSPS, Knoll, Cubbin–Jackson and Douglas). The Emina Scale offers the highest scores for inter-rater reliability and the area under the ROC curve. The Cubbin–Jackson Scale has been used in two validation studies, while the others have been used in just one. Therefore, their validation requires further research for the initial results to be confirmed.

Table 6 Meta-analysis of pressure ulcer risk indicators (random effects model)

Nurses’ clinical judgement

Discussion

We extracted data concerning nurses’ clinical judgement from three studies. The results (Table 4) show moderate sensitivity values (around 50%), variable specificity, and high positive predictive values, which – to a certain extent – are due to the high incidence of PUs in the studies. Inter-rater reliability was not investigated in these studies.

Aggregate results When the results of the various validation studies are aggregated and the weighted mean is obtained (Table 5), we can see that the scale achieving the best scores is the Braden Scale, while the Norton and Waterlow Scales did not perform better than nurses’ clinical judgement. The Waterlow Scale stands out for its high sensitivity (ability to detect most patients developing PUs), but at the cost of very limited specificity (many not-at-risk patients are identified as at-risk patients, so that unnecessary prevention measures are applied).

Risk indicators We used the OR as an indicator of a RAS’s prediction capacity for PU development. The OR scores obtained for the scales under study show great variability. Meta-analysis of the scales with enough studies are shows that the scale which best predicts PU risk is the Braden Scale (see Table 6). Clinical judgement is not a good predictor of PU risk.

n (number N (number Odds 95% confidence of studies)* of patients) ratio interval

Scale

Braden 16 Norton 5 Waterlow 5 Clinical judgement 3

5847 2008 2215 302

4Æ08 2Æ16 2Æ05 1Æ69

2Æ56–6Æ48 1Æ03–4Æ54 1Æ11–3Æ76 0Æ76–3Æ75

*Only studies with calculated odds ratio are included.

Clinical effectiveness We found only three studies of the clinical effectiveness of RASs for PU prevention, and only one of these was a recent clinical trial. Given this lack of studies, we chose for consideration two studies with a quasi-experimental design, despite their methodological limitations. The RAS for PU prevention used in these studies was the Norton Scale, either the original or its modifications. It should be noted that there are no research studies on the clinical effectiveness of other scales, such as Waterlow or Braden. Although the study design of Gunningberg et al. (1999) is good, it proves inadequate for it is not a random study and deals with a very specific type of patient (hip fracture). The prevention measures considered consisted of systematic use of a RAS for PUs. This study concluded that this kind of intervention has no effects in reducing PU prevalence. Bale et al. (1995), who used a prepost design, also investigated use of the MNS as a criterion for providing pressure-reducing support surfaces to a group of patients in a palliative care centre. The authors found a sharp decrease in PU incidence, but this is because these patients used high-quality pressure-reducing surfaces much more frequently. We believe that the reason for this positive effect was not so much the use of the RAS for PU prevention as a more effective use of prevention intervention. Finally, Hodge et al. (1990) studied how the use of the Norton Scale, after a training course, could affect PU incidence and the frequency of prevention interventions.

Table 5 Accumulated analysis of indicators of validity

Scale

n (number of studies)

N (accumulated number of patients)

Sensitivity (%)*

Specificity (%)*

Positive predictive value (%)*

Negative predictive value (%)*

Efficacy (%)*

Braden Norton Waterlow Clinical judgement

20 5 6 3

6443 2008 2246 302

57Æ1 46Æ8 82Æ4 50Æ6

67Æ5 61Æ8 27Æ4 60Æ1

22Æ9 18Æ4 16Æ0 32Æ9

91Æ0 87Æ0 89Æ0 75Æ9

66Æ7 60Æ2 34Æ4 58Æ0

*Weighted average. 106


2006 Blackwell Publishing Ltd, Journal of Advanced Nursing

Integrative literature reviews and meta-analyses

What is already known about this topic • Risk assessment scales for pressure ulcer are considered useful prevention tools, but there is no evidence that their use in clinical practice decreases pressure ulcer incidence. • There is no evidence that some of the risk assessment scales are better than nurses’ clinical judgements. • The Braden Scale and the Norton Scale are generally considered to be the best validated tools.

What this paper adds • Confirmation of the lack of evidence establishing that the use of risk assessment scales decreases pressure ulcer incidence. • The Braden Scale has the best validity and reliability indicators, and has been used in a large number of studies in a wide variety of settings. • The Braden and Norton Scales predict pressure ulcer development risk better than nurses’ clinical judgement, while the Waterlow Scale has good sensitivity but low specificity.

They found a clear increase in the frequency of prevention interventions in the experimental group, which may have been due either to greater nurse awareness of patient risk or to bias resulting from lack of intervention masking. However, the effects on skin lesion development were practically insignificant, probably because some of the prevention interventions performed are nowadays not considered as preventive but rather as inadequate (such as massages or the use of rubber ring cushions) (Bergstrom et al. 1992, GNEAUPP 2000). If we consider all three studies selected globally, we can conclude that there is no evidence that the scales under study are in themselves effective tools for PU incidence reduction. However, one of the studies shows that they are a useful tool for improving the effectiveness of providing pressure-reducing surfaces, and another shows that they are also useful for improving preventive interventions, which are more frequent and implemented earlier.

Braden Scale The Braden Scale has been subjected to the most complete validation, as far as number of studies and different settings (acute hospitals, chronic hospitals, palliative care centres and home care services) are concerned. Both the data from the
2006 Blackwell Publishing Ltd, Journal of Advanced Nursing

Risk assessment scales for pressure ulcer prevention

validation studies and our own analysis of the scale as a risk factor show that the Braden Scale offers the best balance between sensitivity and specificity, and the best effectiveness score. The weighted positive predictive value is 22Æ9%, which we believe to be reasonable considering that it is influenced by the different PU incidences in different studies. A patient scoring below the cut-off score (as established by a particular medical context), that is, a patient at risk of PU, is between 2Æ5 and six times as likely to develop PU as a one scoring higher on this scale. In addition, inter-rater reliability is high, so that consistent results are obtained when using it in different settings.

Norton Scale Although the Norton Scale was the first RAS for PU prevention ever developed, it has been subjected to very few validation studies. Considering all the validation data and its analysis as a risk factor tool, it stands in second position in validity ranking. Its sensitivity and positive predictive value can be considered reasonable. Its main disadvantage is the lack of studies dealing with inter-rater reliability: because the parameters lack operational definitions, application of the scale may give rise to quite different interpretations. In this respect, we think that further validation is needed, including analysis of interrater reliability.

Waterlow Scale The scale developed by Waterlow has good PU risk prediction capacity (OR ¼ 2Æ05 CI 95% ¼ 1Æ11–3Æ76) and high sensitivity, but at the cost of too low specificity. This creates many false positive results, that is, the scale classes as at-risk many patients who are not actually at risk. This means that prevention measures have to be applied to many patients not actually in need of them, which in turn means greater expenditure on prevention materials and more time devoted to nursing work.

Other scales We identified several RASs for PU prevention which have been subjected to only one validation study, so that the evidence is too small to consider their use. First, there is the Cubbin–Jackson Scale, which was specifically designed for patients admitted to intensive care units. The validation indicators of this scale are good, which means that this tool may prove suitable for this type of hospital unit. In order to assess the clinical effectiveness of the scale, it should be 107

P.L. Pancorbo-Hidalgo et al.

subjected to further validation studies in intensive care units with larger samples. According to the only validation study published, the Emina Scale, which was developed from the Norton Scale, also shows good validation and good PU risk prediction indicators (OR ¼ 8Æ24). The scale is easy to apply and has good prediction capacity. Further validation studies would probably confirm its prediction capacity and clinical effectiveness, thereby making it an effective tool to reduce PU incidence.

Clinical judgement Nurses’ clinical judgement has reasonable sensitivity (about 50%) in identifying patients at risk of PU. However, it does not have enough risk prediction capacity, since the 95% confidence interval of the OR includes 1. However, the studies which include nurses’ clinical judgement do not give data on the nurses’ experience or qualifications. We believe this to be important, since the ability to provide a correct clinical judgement is greater among professionally experienced nurses than among inexperienced ones. It is precisely to avoid the problem of lack of experience that RASs for PU prevention have been developed, so that any nurse who knows how to use it may give correct estimates about a patient’s risk of developing PUs. Buhrer and Mitchell (1996) conducted a study of the way nurses who are experienced in PU prevention reach clinical judgements about a patient’s risk. They found that, although they initially used the same factors as those included in the major RASs (Braden, Norton and Waterlow), they weighed each factor, giving greater weight to nutrition and physical activity. What is more, they took into consideration up to 35 different factors not included in these scales but which helped them to gauge the actual risk of each patient. Therefore, assessing PU risk is a complex process implying both a thorough overall assessment and the use of skills resulting from experience, such as intuition and internal evidence.

Systematic reviews on RASs for PU prevention Our own results agree with the systematic reviews conducted by Cullum et al. (1995) and McGough (2000), that is, there is no evidence that RASs for PU prevention, considered independently, are effective. However, our review is at variance with these since their authors studied concluded that none of the RAS for PU prevention considered was better or more effective than clinical judgement, even though the latter highlighted the Braden Scale as the best validated. We believe that this is because the validation studies conducted over the 108

last few years give enough data to confirm the clinical usefulness of some of these scales.

Conclusions There is presently not enough evidence to claim that use of a RAS in clinical practice decreases PU incidence. The use of a validated RAS, the Norton Scale, as a criterion for prevention intervention (pressure-reducing support surfaces) increases both its effectiveness and the application of a greater number of early prevention interventions. We found that the Braden and Norton Scales are better risk prediction tools than nurses’ clinical judgement. The weighted analysis and meta-analysis of the validation studies give enough reasonable evidence for the Braden Scale to be recommended as the one offering the best sensitivity/specificity balance and the highest prediction capacity. There is presently no evidence that nurses’ clinical judgement on its own can predict PU development risk in all patients.

Acknowledgements We would like to acknowledge the contribution as co-reviewer of Andre´s Roldan, University Hospital ‘Virgen del Rocio’. We thank the Foundation Index (Granada, Spain) for the use of bibliographic database CUIDEN. This study was funded by a grant from the Health Institute Carlos III, Ministry of Health and Consumer (Spain) (no. 02/10044).

Author contributions PLP and FGF were involved in the study conception and design. PLP, FGF, IML and CA performed the data collection, data analysis and drafting of the manuscript. PLP obtained funding and supervised.

References Andersen K.E., Jensen O., Kvorning S.A. & Bach E. (1982) Prevention of pressure sores by identifying patient at risk. British Medical Journal 284, 1370–1371. Bale S., Finlay I. & Harding K.G. (1995) Pressure sore prevention in a hospice. Journal of Wound Care 4(10), 465–468. Barnes D. & Payton R.G. (1993) Clinical application of the Braden scale in the acute-care setting. Dermatology Nursing 5(5), 386– 388. Bergquist S. & Frantz R. (2001) Braden Scale: validity in communitybased older adults receiving home health care. Applied Nursing Research 14(1), 36–43. Bergstrom N., Allman R.M., Carlson C.E., Eaglstein W., Frantz R.A., Garber S.L., Gosnell D., Jackson B.S., Kemp M.G., Krouskop T.A., Marvel E.M., Rodeheaver G.T. & Xakellis G.C.
2006 Blackwell Publishing Ltd, Journal of Advanced Nursing

Integrative literature reviews and meta-analyses (1992) Pressure Ulcers in Adults: Prediction and Prevention. Guideline report no. 3. US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research, Rockville, MD. AHCPR Publication no. 93-0013. Bergstrom N., Allman R.M., Alvarez O.M., Bennett M.A., Carlson C.E., Frantz R.A., Garber S.L., Jackson B.S., Kaminski M.V. Jr, Kemp M.G., Krouskop T.A., Lewis V.L. Jr, Maklebust J.A., Margolis D.J., Marvel E.M., Reger S.I., Rodeheaver G.T., Salcido R., Xakellis G.C. & Yardony G.M. (1994) Treatment of Pressure Ulcers. Clinical Practice Guideline no. 15. US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research, Rockville, MD. AHCPR Publication no. 95-0652. Bergstrom N., Braden B., Kemp M., Champagne M. & Ruby E. (1998) Predicting pressure ulcer risk. A multisite study of the predictive validity of the Braden scale. Nursing Research 47(5), 261–269. Bergstrom N., Braden B., Laguzza A. & Holman V. (1987a) The Braden Scale for predicting pressure sore risk. Nursing Research 36(4), 205–210. Bergstrom N., Demuth P.J. & Braden B. (1987b) A clinical trial of the Braden scale for predicting pressure sore risk. Nursing Clinics of North America 22(2), 417–429. Best Practice (1997) Pressure Sores. Part 1: prevention of pressure related damage. The Joanna Briggs Instituted for Evidence Based Nursing and Midwifery 1(1), 1–6. Boyle M. & Green M. (2001) Pressure sores in intensive care: defining their incidence and associated factors and assessing the utility of two pressure sore risk assessment tools. Australian Critical Care 14(1), 24–30. Braden B. & Bergstrom N. (1994) Predictive validity of the Braden scale for pressure sore risk in a nursing home population. Research in Nursing and Health 17, 459–470. Buhrer R. & Mitchell P. (1996) Peer-identified expert nurses’ approaches to risk assessment for pressure ulcers. Advances in Wound Care 9(1), 41–46. Capobianco M.L. & McDonald D.D. (1996) Factors affecting the predictive validity of the Braden scale. Advances in Wound Care 9(6), 32–36. Colin D., Barrois B., Allaert F.A. & Bontoux L. (1997) Ana´lisis epidemiolo´gico y costo de las u´lceras por presio´n. Gerokomos/ Helcos VIII(20), IV–VIII. Critical Appraisal Skills Programme (2002) 10 questions to help you make sense of randomised controlled trials. Retrieved from: http:// www.phru.nhs.uk/casp/casp_rct_tool.pdf on 9 January 2004. Cullum N., Deeks J., Fletcher A., Mouneimne H., Sheldon T., Song F. & Long A. (1995) The prevention and treatment of pressure sores: how useful are the measures for scoring people’s risk of developing a pressure sore? Effective Health Care 2(1), 1–18. Delgado M. (2002) Revisio´n sistema´tica de estudios. Metaana´lisis. UD7. In: Disen˜o de Estudios Sanitarios (Domenech J.M., ed.), Signo, Barcelona. Edwards M. (1995) The levels of reliability and validity of the Waterlow pressure sore risk calculator. Journal of Wound Care 4(8), 373–378. European Pressure Ulcer Advisory Panel (EPUAP) (1998) Pressure ulcer prevention guidelines. EPUAP Review I(1), 7–45.


2006 Blackwell Publishing Ltd, Journal of Advanced Nursing

Risk assessment scales for pressure ulcer prevention Fuentelsaz Gallego C. (2001) Validacio´n de la escala EMINA
: un instrumento de valoracio´n del riesgo de desarrollar u´lceras por presio´n en pacientes hospitalizados. Enfermerı´a clı´nica 11(3), 97– 103. Ga´lvez Toro A. (2001) Enfermerı´a basada en la evidencia. Co´mo incorporar la investigacio´n a la pra´ctica de los cuidados. Fundacio´n Index, Granada. Garcı´a Ferna´ndez F.P., Pancorbo Hidalgo P.L. & Laguna Parras J.M. (2001) Guı´a para el cuidado del paciente con u´lceras por presio´n o con riesgo de desarrollarlas. 1st edn, H.U. Princesa de Espan˜a, Consejerı´a de Salud, Junta de Andalucı´a, Jae´n. Goodridge D.M., Sloan J.A., LeDoyen Y.M., McKenzie J.A., Knight W.E. & Gayari M. (1998) Risk-assessment scores, prevention strategies, and the incidence of pressure ulcers among the elderly in four canadian health-care facilities. Canadian Journal of Nursing 30(2), 23–34. GNEAUPP, Grupo Nacional para el Estudio y Asesoramiento en Ulceras por Presio´n y Heridas Cro´nicas (2000) Documentos GNEAUPP. Ed. GNEAUPP, Logron˜o, Spain. Gunningberg L., Lindholm C., Carlsson M. & Sjo¨de´n P.O. (1999) Implementation of risk assessment and classification of pressure ulcers as quality indicator for patients with hip fractures. Journal of Clinical Nursing 8, 396–406. Hagisawa S. & Barbenel J. (1999) The limits of pressure sore prevention. Journal of the Royal Society of Medicine 92, 576– 578. Halfens R.J.G., Achterberg T. & Bal R.M. (2000) Validity and reliability of the Braden scale and the influence of others risk factors: a multi-centre prospective study. International Journal of Nursing Studies 37, 313–319. Hodge J., Mounter J., Gardner G. & Rowley G. (1990) Clinical trial of the Norton scale in acute care settings. The Australian Journal of Advanced Nursing 8(1), 39–48. Land L. (1995) A review of pressure damage prevention strategies. Journal of Advanced Nursing 22, 329–337. Langemo D.K., Olson B., Hunter S., Hanson D., Burd C. & Cathcart-Silberberg T. (1991) Incidence and prediction of pressure ulcers in five patient care setting. Decubitus 4(3), 25–33. Lewicki L.J., Mion L.C. & Secic M. (2000) Sensitivity and specificity of the Braden scale in the cardiac surgical population. Journal of Wound Ostomy Continence Nursing 27(1), 36–41. Lindgren M., Unosson M., Krantz A.M. & Ek A.C. (2002) A risk assessment scale for the prediction of pressure sore development: reliability and validity. Journal of Advanced Nursing 38(2), 190– 199. Lowthian P. (1989) Identifying and protecting patients who may get pressure sores. Nursing Standard 4(4), 26–29. Lyder C.H., Yu C., Emerling J., Mangat R., Stevenson D., EmpleoFrazier O. & McKay J. (1999) The Braden Scale for pressure ulcer risk: evaluating the predictive validity in black and latino/hispanic elders. Applied Nursing Research 12(2), 60–68. Lyder C.H. (2002) Pressure ulcer prevention and management. Annual Review Nursing Research 20, 35–61. Lyder C.H. (2003) El problema de las UPP en los Estados Unidos: epidemiologı´a, coste, consideraciones legales y aspectos e´ticos. Retrieved from http://multimedia.cesanitaris.com/gneaupp2003/ lyder.pdf on 9 January 2004.

109

P.L. Pancorbo-Hidalgo et al. Maklebust J.A. & Sieggreen M. (2001) Pressure Ulcers. Guideline for Prevention and Management, 3rd edn. Springhouse Corporation, Springhouse, PA. McGough A. (2000) A systematic review of the effectiveness of risk assessment scales used in the prevention and management of pressure sores. In Pressure Ulcer Risk Assessment and Prevention (Rycroft-Malone J. & McInnes E., eds), technical report. Royal College of Nursing, London, pp. 84–99. Pancorbo Hidalgo P.L. & Garcı´a Ferna´ndez F.P. (2002) Estimacio´n del coste econo´mico de la prevencio´n de u´lceras por presio´n en una unidad hospitalaria. Gerokomos 13(3), 164–171. Pang S.M. & Wong T.K. (1998) Predicting pressure sore risk with the Norton, Braden and Waterlow scales in a Hong Kong rehabilitation hospital. Nursing Research 47(3), 147–153. Perneger T.V., Rae¨ A.C., Gaspoz J.M., Borst F., Vitek O. & He´liot C. (2002) Screening for pressure ulcer risk in an acute care hospital: development of a brief bedside scale. Journal of Clinical Epidemiology 55(5), 498–504. Ramundo J.M. (1995) Reliability and validity of the Braden scale in the home care setting. Journal of Wound Ostomy and Continence Nursing 22(3), 128–134. Research Dissemination Core (2002) Prevention of Pressure Sore. Gerontological Nursing Interventions Research Center, University of Iowa, Iowa City. Rycroft-Malone J. & McInness E. (2000) Pressure Ulcer Risk Assessment and Prevention. Technical Report. RCN, London. Sackett D.L., Richardson W.S., Rosenberg W. & Haynes R.B. (1997) Evidence-Based Medicine: How to Practice and Teach EBM. Churchill Livingstone, London. Salvadalena G.D., Snyder M.L. & Brogdon K.E. (1992) Clinical trial of the Braden Scale on an acute care medical unit. Journal of Enterostomal Therapy Nursing 19(5), 160–165. Schoonhoven L., Haalboom J.R., Bousema M.T., Algra A., Grobbee D.E., Grypdonck M. & Buskens E. (2002) Prospective cohort study of routine use of risk assessment scales for prediction of pressure ulcers. British Medical Journal 325(12), 797– 800. Seongsook J., Ihnsook J. & Younghee L. (2004) Validity of pressure ulcer risk assessment scale: Cubbin and Jackson, Braden, and Douglas scales. International Journal of Nursing Studies 41(2), 199–204. Smith D.M. (1995) Pressure ulcers in the nursing home. Annals of Internal Medicine 123, 433–442.

110

Smith I. (1989) Waterlow/Norton scoring system A ward view. CARE-Science and Practice 7(4), 93–95. Smith L.N., Booth N., Douglas D., Robertson W.R., Walker A., Durie M., Frasier A., Hillan E.H. & Swaffield J. (1995) A critique of ‘at risk’ pressure sore assessment tools. Journal of Clinical Nursing 4, 153–159. Soldevilla Agreda J.J. & Torra i Bou J.E. (1999) Epidemiologı´a de las u´lceras por presio´n en Espan˜a. Estudio piloto en la Comunidad Auto´noma de La Rioja. Gerokomos/Helcos 10(2), 75–87. Stotts N.A. (1988) Predicting pressure ulcer development in surgical patients. Heart and Lung 17, 641–647. The Nursing Best Practice Guideline (2003) Risk Assessment and Prevention of Pressure Ulcers. Registered Nurses’ Association of Ontario, Ontario. Retrieved from http://www.rnao.org on 30 June 2003. Thomas D.R. (2001) Issues and dilemmas in the prevention and treatment of pressure ulcers: a review. Journal of Gerontology: Medical Sciences 56A(6), M328–M340. Thoroddsen A. (1999) Pressure sore prevalence: a national survey. Journal of Clinical Nursing 8, 170–179. Torra i Bou J.E., Rueda J., Soldevilla J.J., Martı´nez F. & Verdu´ J. (2003) 1er Estudio Nacional de Prevalencia de u´lceras por presio´n en Espan˜a. Epidemiologı´a y variables definitorias de las lesiones y pacientes. Gerokomos 14(1), 37–47. Torra i Bou J.E. (1997) Valorar el riesgo de presentar u´lceras por presio´n. Escala de Braden. Revista Rol de Enfermerı´a 20(224), 23–30. Towey A.P. & Erland S.M. (1988) Validity and reliability of an assessment tool for pressure ulcer risk. Decubitus 1(2), 40–48. VandeBosch T., Montoye C., Satwicz M., Durkee-Leonard K. & Boyland-Lewis B. (1996) Predictive validity of the Braden scale and nurse perception in identifiying pressure ulcer risk. Applied Nursing Research 9(2), 80–86. Verdu´ J., Nolasco A. & Garcı´a C. (2003) Analisis y evolucio´n de la mortalidad por u´lceras por presio´n en Espan˜a. Periodo 1987–1999. Gerokomos 14(4), 212–226. Wai-Han C., Kit-Wai C., French P., Yim-Sheung L. & Lai-Kwan T. (1997) Which pressure sore risk calculator? A study of the effectiveness of the Norton scale in Hong Kong. International Journal of Nursing Studies 34(2), 165–169. Westrate J.T.M., Hop W.C.J., Aalbers A.G.J., Vreeling A.W.J. & Bruining H.A. (1998) The clinical relevance of the Waterlow pressure sore risk scale in the ICU. Intensive Care Medicine 24, 815–820.


2006 Blackwell Publishing Ltd, Journal of Advanced Nursing