VOL: 103, ISSUE: 14, PAGE NO: 32-33
Jane Willock, RGN, RSCN, BSc, PGDipEd, MSc; Mona Baharestani, PhD, CWOCN, FCCWS, FAPWCA, is director of wound healing, Long Island Jewish ; Denis Anthony, RMN, RGN, RN (Canada), BA, MSc, PhD
Jane Willock is senior lecturer, faculty of health and sports sciences, University of Glamorgan, and nurse practitioner, children?s investigations unit, University Hospital of Wales, Cardiff; Dr Mona Baharestani is director of wound healing, Long Island Jewish Medical Centre, New York; Professor Denis Anthony, is head of postgraduate research studies, faculty of health and life sciences, De Montfort University, Leicester.
Abstract: Willock, J. et al (2007) A risk assessment scale for pressure ulcers in children. nursingtimes.net.
Abstract:Willock, J. et al (2007) A risk assessment scale for pressure ulcers in children. nursingtimes.net. Aim:To develop a pressure ulcer risk assessment scale for children using statistical methods and patient data. Method:Data on 336 children admitted to 11 hospitals was collected using questionnaires. The data was studied to compare the characteristics of children who had developed pressure ulcers with the characteristics of the representative sample of hospitalised children. The significance of the children’s characteristics in the development of pressure ulcers was then estimated and a risk assessment scale developed. Results:Using the significance values as a guide, the Glamorgan Paediatric Pressure Ulcer Risk Assessment Scale was developed. High weightings were assigned to immobility and pressure on the skin, but a lower weighting was given to anaemia. At a risk score of 10 the scale was 100% sensitive but 50.2% specific. It is difficult to assess true sensitivity and specificity due to the impact of preventive measures. At a risk score of 15, the scale was 98.4% sensitive and 67.4% specific. The area under the ROC curve was found to be 0.912, giving the Glamorgan scale a predictive validity of 91.2%. Discussion:The Glamorgan scale appears to be the first paediatric pressure ulcer risk assessment scale developed statistically using patient data. The sensitivity, specificity and predictive validity of the Glamorgan scale appears to be greater than the Braden Q scale, but testing with new data sets is required. Conclusion:The Glamorgan scale may give a more accurate estimate of risk than other scales, but it is important to note that no risk assessment scale can be 100% accurate. Nurses should examine and try to resolve the individual problems that contribute to the total risk. Pressure ulcer risk assessment scales are used to identify patient risk, level of risk and type of risk. While more than 200 published pressure ulcer risk factors have been cited, (Salzberg et al, 1999), not all characteristics are relevant to all patient groups, and clearly it would be impossible and impractical to incorporate more than a few risk factors into an assessment scale. Adult pressure ulcer risk assessment scales were initially developed based on patient observation (Norton et al, 1962; Waterlow, 1985) and later on literature reviews (Braden and Bergstrom, 1987), which identified factors believed to predispose patients to pressure ulcer development. Although some risk assessments have undergone extensive testing, the individual predictive value of each item was not taken into account, and nor were comprehensive analyses of all variables performed (Haalboom et al, 1999). The prevalence of pressure ulcers in hospitalised children has been estimated between 0.47% and 13.1% (Baldwin, 2002; Groeneveld et al, 2004; Willock et al, 2000; Waterlow, 1997; McLane et al, 2004; Dixon and Ratcliff, 2005) and up to 27% in paediatric intensive care units (Zollo et al, 1996; Curley et al, 2003; Curley et al, 2000). However, there are very few research publications describing the characteristics of children with pressure ulcers (McCord et al, 2004; Willock et al, 2005; Willock et al, 2000). Ten published paediatric risk assessment scales have been identified, of which six are modifications of adult risk assessment scales (Quigley and Curley, 1996; Garvin, 1997; Huffines and Logsdon, 1997; Pickersgill, 1997; Samaniego, 2003; Suddaby et al, 2005). Two risk assessments are based on patient observation (Bedi, 1993; Olding and Patterson, 1998), one on a review of relevant literature (Cockett, 1998), and one on a multi-centre survey but this was not predictive (Waterlow, 1997; Waterlow, 1998). No published studies were identified using statistical methods to develop a pressure ulcer risk assessment scale directly from patient data. MethodAn initial survey of 265 patients (seven of whom had pressure ulcers) in a children’s hospital in England was performed to obtain detailed patient characteristic data representative of a paediatric hospitalised population. Detailed questionnaires were developed based on a review of the paediatric and adult pressure ulcer literature and extensive discussions with paediatric nurses experienced in pressure ulcer prevention and care. Using these questionnaires, data on the characteristics of 336 paediatric inpatients in 11 hospitals was collected as described in Table 1 (Willock et al, 2000; Willock et al, 2005). Table 1. Details of the sample of children
|Sites||Royal Liverpool Children’s Hospital||Eleven hospitals in England and Wales*|
|Data collection type||Prevalence study - all inpatients during one day||Incidence study in three wards over one month||Survey of children who developed pressure ulcers (and some with similar characteristics who did not develop pressure ulcers)|
|Sample size||With pressure ulcer||4||3||54|
|Without pressure ulcer||179||79||17|
|Sample characteristics||Mainly Caucasian, aged from one day to 17 years 11 months, 60% boys||Mainly Caucasian, aged three weeks to 17 years and eight months, 61% boys|
* The 11 hospitals inEnglandandWaleswere:
A prevalence study measures the proportion of cases in a population at a point in time, and an incidence study measures the proportion of new cases presenting over a period of time. This sort of data is more representative of the general hospitalised paediatric population than an unstructured survey, and can be used to compare with data from a population of children specifically with pressure ulcers. The European Pressure Ulcer Advisory Panel’s (EPUAP, 2005) and the National Pressure Ulcer Advisory Panel’s (NPUAP, 2003) pressure ulcer staging systems were used. Local ethical approval was obtained for the single site study, and multi-centre ethics approval was obtained for the study in 11 hospitals. Data collectors in both studies received standardised training on pressure ulcer identification and questionnaire administration. Data collected from both studies was combined so that the characteristics of children who had developed pressure ulcers could be compared with the characteristics of the representative sample of hospitalised children. As the data were predominantly nominal, chi-square tests were performed using SPSS (Statistical Package for the Social Sciences version 12) to estimate the significance of the children’s characteristics in the development of pressure ulcers. ResultsTable 2. The level of significance of variables associated with pressure ulcers in children
Level of significance
|Difficult to position|
|Equipment pressing or rubbing|
|Poor peripheral perfusion|
|Low serum albumin|
|Weight below 10 thcentile for age|
|Incontinence inappropriate for age||.003|
|Poor tissue oxygenation||.194|
|Reduced conscious level||.236|
|Weight over the 90 thcentile for age||.275|
|Self care ability inappropriate for age||.465|
Using the significance values presented in table 2 as a guide, the Glamorgan Paediatric Pressure Ulcer Risk Assessment Scale (the Glamorgan scale) was developed. While other paediatric risk assessment scales give similar weightings to each subscale, our analyses have found that some variables, such as mobility, are more significant than others, and should be accordingly weighted. High weightings were assigned to ‘immobility’ and ‘pressure on the skin’, as all of the children with pressure ulcers showed at least one of these characteristics. The item ‘difficult to position’ was considered to be a characteristic of mobility. Illustratively, critically ill patients may be difficult to position or reposition if they have decreased oxygen saturation levels and hypotension when they are moved. Risk scores were adjusted so that the higher the total score, the greater the pressure ulcer risk (10+ = at risk; 15+ = high risk; 20+ = very high risk). Although anaemia emerged as being highly significant, all of the children with pressure ulcers were either immobile or had equipment pressing on their skin. None of the children who were mobile and did not have equipment pressing on their skin developed pressure ulcers, even if they were anaemic. Therefore it was decided by the authors (who have experience in pressure ulcer prevention) that anaemia should have a lower weighting than immobility. Content validityThe content validity of the Glamorgan scale was established by testing as many variables as possible during the initial data collection period and rejecting all the variables that did not reach statistical significance. Sensitivity and specificitySensitivity is the degree to which a scale accurately identifies all those people who are at risk of developing a pressure ulcer. Specificity is the degree to which a scale will correctly identify those not at risk of pressure ulcer development. Sensitivity and specificity are difficult to capture, as the risk of developing a pressure ulcer is not the same as actually developing a pressure ulcer, given that carers actively intervene to prevent pressure damage. The data set used to develop the Glamorgan scale was also used to calculate the sensitivity and specificity. No child with a score of less than 10 on the Glamorgan scale developed a pressure ulcer. All children scoring 10 or greater were assessed as being at risk of developing a pressure ulcer. Those scoring 9 or below were assessed as not at risk. At a risk score of 10 the risk assessment scale was found to be 100% sensitive (none of the children with a score of less than 10 developed a pressure ulcer), but only 50.2% specific (50.2% of children who did not develop a pressure ulcer were correctly predicted not to be at risk). Some 49.8% of children who did not develop a pressure ulcer were predicted to be at risk (false positives). However, some of these children may actually have been at risk of developing a pressure ulcer but intervention prevented tissue damage. Risk scores will appear to be performing poorly if preventive care is very effective. To determine true sensitivity and specificity this would require the unethical withholding of preventive intervention (Flanagan, 1995). At a risk score of 15, the Glamorgan scale is 98.4% sensitive and 67.4% specific. However, if a score of 15 is used as a cut-off, some children who are at risk of developing pressure ulcers may be misdiagnosed and preventative action may not be taken. Predictive validityTo test the predictive validity of the Glamorgan scale, the area under the receiver operating characteristic (ROC) curve was calculated. An ROC curve is a plot of the true positive rate (sensitivity) against the false positive rate (1 - specificity) for given score weightings or thresholds (Anthony et al, 2003). The area under the curve is calculated and can indicate whether any randomly selected patient at risk of developing a pressure ulcer will have a higher score than any patient not at risk of developing a pressure ulcer (Anthony, 1996). Fig 1 is an ROC curve using the Glamorgan scale. The area under the curve calculated by SPSS is 0.912, indicating that for any patient there is a 91.2% chance that the Glamorgan scale will correctly predict their risk of developing a pressure ulcer. If a risk assessment could be designed that had an area under the curve of 1.0, it would predict risk accurately every time, so the closer to 1.0 the area under the curve is, the better at predicting risk the scale is. If the scale did not have any predictive value, the area under the curve would be 0.5. Criterion validityThere is little published information on the validity of paediatric pressure ulcer risk assessment scales, and no gold standard in adult or paediatric pressure ulcer risk assessment exists. The Braden Q scale (Quigley and Curley, 1996) is a modification of the adult Braden scale and is widely used in the US to assess pressure ulcer risk in children aged 21 days to eight years of age (Baharestani et al, 2007). The data set used to develop and test the Glamorgan scale was also used to calculate the sensitivity, specificity and area under the ROC curve of the Braden Q scale. Unlike most risk assessment scales, the Q scale has low scores for high risk and high scores for low risk. The only risk score at which it is 100% sensitive (a score of 29), it is 0% specific - this means it only correctly identifies all children who will develop pressure ulcers when it assumes all children are at risk irrespective of their characteristics. The highest sensitivity and specificity it achieves is a sensitivity of 67.2% and a specificity of 64.8% at a risk score of 21. The area under the ROC curve was 0.697 (calculated using SPSS), so for any patient there is a 69.7% chance that the Braden Q scale will correctly predict their risk of developing a pressure ulcer (Fig 2). Using this data, the Glamorgan scale demonstrates higher sensitivity and specificity, and better predictive value than the Braden Q scale. DiscussionAs far as possible, patient care should be evidence-based. Although this study shows that it is possible to collect evidence with which to develop pressure ulcer risk assessment scales, most scales have been developed in an ad hoc fashion based on opinions of the relative importance of possible risk factors (EHCB, 1995). All of the paediatric pressure ulcer risk assessment scales described in the literature have been either derived from adult risk assessment scales, clinical observation or literature review. If a risk assessment scale is used it is recommended that it has been tested for use in the specialty in which it is used (RCN, 2001). It is therefore not acceptable simply to just modify an adult scale for use with children. The Glamorgan scale appears to be the first paediatric pressure ulcer risk assessment scale developed statistically using patient data. While acknowledging that the sample size used to develop the Glamorgan scale was not large, every attempt was made to include data representative of the hospitalised paediatric population. While the sensitivity, specificity and predictive validity of the Glamorgan scale appears to be greater than the Braden Q scale, these calculations are based on data used to develop the Glamorgan scale. In order to test the inter-rater reliability and validity of the Glamorgan scale and compare it with other paediatric risk assessment scales, testing with new data sets is required. To be effective, risk assessment scales must identify those at risk, their level of risk and the specific factors placing them at risk. Early risk identification should be directly linked to a risk-based prevention plan, where evidence-based resource allocation is informed by the risk assessment scale together with clinical expertise. The greatest clinical use of risk assessment scales may be in the development of pressure ulcer prevention protocols, which combine assessments of risk factors with recommendations for care and further surveillance (Deeks, 1996). Application to practiceMost pressure ulcer risk assessment scales give similar weightings to each subscale, such as nutrition, continence and mobility. However, pressure ulcers are caused by prolonged pressure on the skin, shearing or friction when the skin rubs against a surface. In order to prevent pressure ulcers in children, these are the problems nurses need to monitor and take preventative action. Other characteristics, such as poor nutritional intake and low serum albumin, may increase a child’s risk of developing a pressure ulcer, but without pressure, shear force or friction, they will not lead to pressure ulcer formation. However, to reduce risk, these characteristics should be resolved as much as possible. It is important not to think of pressure ulcer risk as a single score, but to look at and try to resolve the individual problems that contribute to the total risk. The Glamorgan scale may give a more accurate estimate of pressure ulcer risk than other scales, but it is important to remember that no pressure ulcer risk assessment scale can be 100% accurate. ReferencesAnthony, D. et al(2003) A regression analysis of the Waterlow score in pressure ulcer risk assessment. Clinical Rehabilitation; 17: 2, 216-223. Anthony, D.(1996)Issues in research. Receiver operating characteristic analysis: an overview.Nurse Researcher; 4: 2, 75-88. Baharestani, M.M. et al(2007). Pressure ulcers in neonates & children: An NPUAP White Paper. Advances in Skin and Wound Care, January 2007. Baldwin, K.M.(2002) Incidence and prevalence of pressure ulcers in children. Advances in Skin and Wound Car e; 15: 3, 121-124. Bedi, A.(1993) A tool to fill the gap: developing a wound risk assessment chart for children. Professional Nurse; 9: 2, 112-120. Braden, B.,Bergstrom N(1987) A conceptual schema for the study of the etiology of pressure sores.Rehabilitation Nursing; 12: 1, 8-12. Cockett, A.(1998) Paediatric pressure sore risk assessment. Journal of Tissue Viability; 8: 1, 30. Curley, M.A.Q. et al(2003) Predicting pressure ulcer risk in pediatric patients - the Braden Q scale. Nursing Research; 52: 1, 22-31. Curley, M.A.Q. et al(2000) The effects of early and repeated prone positioning in pediatric patients with acute lung injury. Chest; 118: 1, 156-163. Deeks, J.(1996) Pressure sore prevention: using and evaluating risk assessment tools. British Journal of Nursing; 5: 5, 313-320. Dixon, M., Ratcliff, C.(2005) Pediatric pressure ulcer prevalence - one hospital’s experience. Ostomy/Wound management; 51: 6, 44-50. EHCB(1995) NHS Centre for Reviews and Dissemination. The prevention and treatment of pressure sores. Effective Health Care Bulletin; 2: 1. EPUAP(2005) European Pressure Ulcer Advisory Panel, EPUAP Statement, Pressure Ulcer Classification: Differentiation Between Pressure Ulcers and Moisture Lesions: http://www.epuap.org/review6_3/page6.html Accessed 8.12.06Flanagan, M.(1995) Who is at risk of a pressure sore? A practical review of risk assessment systems. Professional Nurse; 10: 5, 305-308. Garvin, G.(1997) Wound and skin care for the PICU. Critical Care Nursing Quarterly; 20: 1, 62-71. Groeneveld, A. et al(2004) The prevalence of pressure ulcers in a tertiary care pediatric and adult hospital. Journal of WOCN; 31: 3, 108-122. Haalboom, J.R.E et al(1999) Risk assessment tools in the prevention of pressure ulcers. Ostomy/Wound Management; 45: 2, 20-34. Huffines, B., Logsdon, M.C.(1997) The neonatal skin risk assessment scale for predicting skin breakdown in neonates. Issues in Comprehensive Pediatric Nursing; 20: 2, 103-114. McCord, S. et al(2004) Risk factors associated with pressure ulcers in the pediatric intensive care unit. Journal of WOCN; 31: 4, 179-183. McLane, K.M. et al(2004) The 2003 national pediatric pressure ulcer and skin breakdown prevalence survey: a multisite study. Journal of WOCN; 31: 4, 168-178. NPUAP(2003) National Pressure Ulcer Advisory Panel Staging Report: http://www.npuap.org/positn6.htmlAccessed 8.12.06 Norton, D. et al(1962) An Investigation of Geriatric Nursing Problems in Hospital. Edinburgh: Churchill Livingstone. Olding, L., Patterson, J.(1998) Growing concern, Nursing Times; 94: 38, 74-79. Pickersgill, J.(1997) Taking the pressure off. Paediatric Nursing; 9: 8, 25-27. Quigley, S.M., Curley, M.A.Q.(1996) Skin integrity in the pediatric population: Preventing and managing pressure ulcers. Journal of the Society of Pediatric Nurses;1: 1, 7-18. RCN(2001) Pressure Ulcer Risk Assessment and Prevention, Clinical Practice Guidelines. London: Royal College of Nursing. Salzberg,C.A.et al(1999) Predicting pressure ulcers during initial hospitalization for acute spinal cord injury.Wounds: A Compendium of Clinical Research and Practice; 11: 2, 45-57. Samaniego,I.A.(2003) A sore spot in pediatrics: risk factors for pressure ulcers, Pediatric Nursing; 29: 4, 278-82. Suddaby, E.C. et al(2005) Practice applications of research. Skin breakdown in acute care pediatrics, Pediatric Nursing; 31: 2, 132-8, 148. Waterlow, J.(1997) Pressure sore risk assessment in children. Paediatric Nursing; 9: 6, 21-24. Waterlow, J.(1998) Pressure sores in children: risk assessment. Paediatric Nursing; 10: 4, 22-23. Waterlow, J.(1985) Pressure sores: a risk assessment card.Nursing Times; 81: 48, 49-55. Willock, J. et al(2005) Identifying the characteristics of children with pressure ulcers: results of a multi-centre survey. Nursing Times;101: 11, 40-3. Willock J et al(2000) Pressure sores in children - the acute hospital perspective. Journal of Tissue Viability, 10: 2, 59-62. Zollo, M.B. et al(1996) Altered skin integrity in children admitted to a pediatric intensive care unit. Journal of Nursing Care Quality; 11: 2, 62-67.