Developing a Pressure Ulcer Risk Assessment Scale for Patients in Long-Term Care

    Pressure ulcers are defined as areas of localized damage to the skin and underlying tissue caused by pressure, shear, or friction.^1,2 They are most common among patients who are immobile, elderly, have no sensation in some area of their body, or in intensive care. Commonly known factors that increase the risk for developing pressure ulcers include immobility, circulatory problems, infections, incontinence, passivity, and decrease in consciousness.^3-6 Sometimes pressure ulcers cause intolerable suffering for the patient. They often are relapsing, painful, and represent a risk for secondary infection. They may affect activities of daily living and social relations.^7,8

    Data published by the Agency for Health Policy Research (AHCPR) show that the prevalence of pressure ulcers in hospitals varies from 4% to 69% and the incidence in hospital settings ranges from 3% to 30%.⁹ Incidence and prevalence vary from one study to another because a different criteria for existing pressure ulcers may be used and a common pressure ulcer grading system does not exist.^10,11 However, the numbers are higher in long-term care wards than they are in acute care settings.^5,12,13 For example, year 2000 statistics from Finland indicate that each year pressure ulcers result in approximately 20,000 days in hospital. Assuming the average expenses per day in long-term and acute care are 130 to 170 EURO ($156 to $204 million US), an estimated 3 million EURO ($3.6 million US) was spent in the care of pressure ulcer patients in 2000.¹⁴

    Increased efficiency and quality requirements in modern healthcare require the prevention and treatment of pressure ulcers and the identification of patients at high risk for their development. According to some indications, present risk assessment scales do not lead to efficient use and allocation of preventive measures in patient care.¹⁵ This article describes the process involved in developing a pressure ulcer risk assessment instrument for use in long-term care. The new instrument is based on available research and 21 existing risk assessment scales with the addition of elements relating to preventive devices and organizational factors. Developing a valid and reliable assessment tool is a long process requiring utilization of scientifically constructed and tested instruments.

Literature Review

    The process of developing the new pressure ulcer risk assessment instrument started with a search for relevant sources published between 1966 and February 2000 and indexed on MEDLINE© using the terms risk assessment and pressure ulcer; this produced 131 sources. A subsequent search for the same time period used the key words scale* or score* and pressure and ulcer* or sore* These searches and subsequent efforts yielded 16 risk assessment scales. An additional five scales were referenced within some of the publications: Shannon’s scale,¹⁶ NPRU,¹⁷ Birty’s PARA,¹⁸ the Sunderland score,¹⁹ and the Stratheden pressure sore risk scale.²⁰ All scales, both tested and nontested for reliability and validity, were reviewed because so few tested scales were found.

    The oldest pressure ulcer risk assessment scale is the Norton scale,²¹ which was created for use with elderly patients in the early 1960s. In the 1970s and 1980s, more scientists became interested in wound risk assessment, leading to creation of the Gossnel,²² Knoll,²³ Braden,²⁴ and Waterlow instruments,²⁵ to name a few. A common phenomenon was that the later the scale was developed, the more risk factors it included. However, many of these scales are based on opinions about risk factors rather than on research evidence.² Some risk assessment scales are said to be based on personal experience,^26,27 personal studies,²⁸ or discussions with colleagues,^24,25 as well as earlier studies.

    A scale based on earlier risk assessment tools may rely on undefined concepts and self-evident risk factors. Only the Braden scale’s concepts have been defined and operationalized.²⁴ Stotts²⁹ discusses operationalizing Norton’s scale concepts in greater detail and Gosnell²² elaborates on and defines the important factors in her own main categories.

    The accuracy of unevaluated scales is difficult to estimate.³⁰ Thus far, few risk assessment scales have undergone validity and reliability testing. The Norton (considered the mother of modern scales³¹ and the Waterlow scales have been tested and found to be reliable according to sensitivity and specificity assessments; the Braden scale is the most studied risk assessment scale.^21,24,25 Comparing the results of these studies is difficult because methods and materials are different and the pressure ulcer classifications used are seldom described. Also, the scales do not take into account the structure of the healthcare organization or the use of preventive devices that have been developed. In Finland, surveys reflect that the use of risk assessment scales is virtually nonexistent.^32,33

    A review of the 21 risk assessment scales found in the literature showed the most commonly addressed concepts are incontinence or continence (considered in 20 of the scales evaluated), mobility (18 scales), nutrition (16 scales), mental status (16 scales), and activity (12 scales). These five concepts (main categories) usually are divided into different components (subcategories) to yield the risk score. For example, in the Norton scale, the concept activity is divided into four subcategories: ambulant, walk/help, chair fast, and bedfast (4, 3, 2, and 1 point(s), respectively). The higher the patient’s score, the greater the risk for developing pressure ulcers. All five concepts are included in five scales: Gosnell,^22,34 Shannon,¹⁶ Knoll,²³ Medley,²⁷ and Ek^35,36 (modification of Norton). About half of the risk assessment scales included the following risk factors: skin type in risk areas, skin condition or skin wounds, predisposing diseases (especially diabetes), physical condition or general health state, vital signs, and hemodynamics. Abnormal blood count, supportive treatments (cytostatics, steroids, or inotropes), and age also were taken into account; whereas, temperature, pain, smoking, and trauma or surgery rarely were considered.

    Additional sources used to develop the new instrument for pressure ulcer risk assessment were earlier studies on Finnish pressure ulcer patients and the facilities in which they were treated.^6,31,33 Studies from Finland and elsewhere consistently identify the following common factors that increase the risk of pressure ulcer development: immobility, incontinence, passivity, and reduced consciousness.^3-5 Some studies indicate that pressure ulcer development may be associated with organizational factors, such as the number of staff members, staff structure, length of hospital stay, and the ward’s actual and maximum number of patients.³³

Methods

    Instrument development. Based on published research and study results using existing scales, an instrument for pressure ulcer assessment was developed by Finnish researchers that included questions relating to the following items: 1) patient risk factors: activity, mobility, mental status, appetite, nutrition, urinary incontinence, fecal incontinence, sensory perception, and skin condition; 2) devices and methods used in patient care: technical devices, bed, mattress, mattress pad, sitting cushions, and care considerations (eg, special diet, massage, counseling for patient and relatives on the prevention of pressure ulcers); and 3) number of staff members and their educational background, maximum number of beds, and beds in use. Researchers hypothesized that staff structure — ie, the number and educational background of nursing personnel on wards during the study — might affect pressure ulcer occurrence. The instrument also included demographic data about patient age, sex, height, weight, waist diameter at navel, goal of treatment, diagnoses, and pressure ulcer history (see Table 1).

    Pilot testing. The testing was conducted in a city hospital in southwestern Finland (city population 160,000). All city hospitals in Finland have geriatric units that treat patients who have not been admitted to a nursing home and who are unable to care for themselves at home. The hospitals have both acute and long-term patients; this study included only the long-term care patients. Only registered nurses were included in the pilot test to ensure the sample was as homogenous as possible in terms of educational background. The response rates in different wards ranged from 70% to 100%.

    First phase. The nurse sample of the first phase of pilot testing was gathered from six different wards (43 registered nurses). All nurses answered the instrument questions concerning the same patient (n = 6) in every ward. Two of these wards provided long-term care and four handled internal medicine patients, some of whom were waiting for transfer to a long-term care unit.

    In the first phase, registered nurses answered the first and second question domains of the risk assessment instrument. Head nurses of six wards answered the third question domain to determine whether the questions were understandable; resultant data could be extracted from hospital statistics and, thereby, double-checked. In the first phase, the instrument utilized an ordinal scale — response options in the patient risk assessment ranged from 0 to 3 (0 = no pressure ulcer risk to 3 = highest risk). The response options for the use of devices and methods to prevent pressure ulcers ranged from 0 to 2 (0 = used daily, 1 = used sometimes, 2 = not used).

    Second phase. The second phase initially included 70 expert nurses who were participating in a 9-month tissue viability specialization course. At the time they received the questionnaires, 20 had attended the course for 8 months and 50 had attended for 4 months. The mean age of the latter group was 39.2 years. The majority of the 50 nurses (64%) were registered nurses and the remaining segment (36%) were specialized nurses, including one assistant head nurse and two head nurses. The nurses had been practicing their profession an average of 11.9 years. Almost half of the nurses who replied (46%) were in charge of wound care in their units. Most of them indicated they treated pressure ulcer patients daily or weekly (52%). They felt they were able to assess their patients’ risks for developing pressure ulcers either reasonably well (48%) or very well (48%); two of the nurses (4%) described their competence in this regard as excellent. Only six nurses (12%) indicated their units used risk assessment scales, most commonly the Braden (two nurses) or the Norton (two nurses) scales. Of the 64 forms (91%) returned, 14 had to be excluded due to incomplete data; 50 (71%) usable questionnaires were available for data analysis.

    The second phase instrument (the expert analysis) included only the first and second question domains. This format was presented to 50 expert nurses who assessed the relevance and clarity of questions with respect to pressure ulcer risk relevance, potential ambiguity, nurses’ ability to define the subcategories, and devices and methods for pressure ulcer prevention. In this phase, yes/no answers were sought — eg, Is this question ambiguous (yes/no).

    Data collection and analysis. In the first phase, the head nurses indicated how many risk assessment scale forms they needed. The data were collected within 1 week between March and April 2001. The researcher trained the head and supervisory nurses on questionnaire completion. A training session also was arranged on each ward the day before the pilot test for the nurses who were on duty to gain information on the comprehensibility of the questions — ie, whether the questions were understandable and the instrument measured what it was designed to measure. Ward nurses chose the patient who was to be included in the pilot test: the only criteria were that the person could not have a pressure ulcer and that he/she had been identified as a long-term care patient. The researcher collected the completed instrument forms in each ward.
For the second phase in May 2001, questionnaires were sent by mail to the tissue viability course instructor to be distributed to her students (nurses trained to take care of wounds). The instructor received questionnaire administration instructions by phone and distributed the instrument to the students during class. An introduction letter was attached to every questionnaire explaining each question. The students returned the completed forms to the instructor at the end of the class and the forms were mailed to the researcher. The instructor did not complete the questionnaire.

    Descriptive statistics (ie, minimum and maximum values, mode, frequencies, and percentage distributions) were used for the first phase to analyze results. Percentages of agreement were computed for both phases. The results were considered good when the agreement was greater than 70%.

Results

    Phase one. In the first phase, 100% agreement was reached regarding assessment of urinary incontinence; some discrepancy was noted between days if patient incontinent status changed because of catheter use. Responses regarding activity, mental status, and nutrition assessment achieved 100% agreement on two wards; the other areas of consideration achieved 100% agreement on one ward. The only risk factor for which 100% agreement was never reached was mobility. The highest average agreement (87%) was recorded for the sensory perception assessment. Mean agreement was less than 70% for mobility (in five wards) and appetite (in three wards). Of note: on ward 5, the agreement score for urinary incontinence also was low (43%) (see Table 2).

    For questions regarding devices and methods to prevent pressure ulcers, 46 items were divided into six categories: technical devices, special beds, mattresses, mattress pads, cushions, and care methods. The levels of agreement regarding the use of items were categorized 100%, at least 70%, or less than 70%. When the respondents were asked whether the device or method was used daily, occasionally, or not at all, agreement levels of less than 70% were recorded primarily regarding positioning pillows (eg, under the calf) (three wards); transfer boards (three wards); electronic beds (three wards); patient counseling (three wards); and patient’s significant others’ counseling (four wards). All six wards demonstrated at least 70% agreement with regard to using or not using genuine sheepskin and hydraulic hoisting devices for daily nursing care of patients. Regarding care methods, at least 70% agreement was recorded for the use of a basic cream applied to the skin in areas susceptible to pressure and continuous monitoring of the skin. Mattresses, mattress pads, and seating cushions were evaluated in a single unit and appeared to be never used by the nurses in the study.

    Using the results from the first phase, the instrument was revised. The main category “appetite” was deleted. The words “in bed” were added to the main category “mobility” as well as its subcategories. Some clarifying examples also were added to the subcategories. For instance, an example about a wheelchair patient was added to the activity’s first and third subcategories. The second subcategory under “nutrition” was explained by including the statement, “can eat at least half the portion by him/herself”. Also, the first subcategory “continent” under the main category “urinary incontinence” was clarified by adding “eg, catheter patient”. No changes were made to either the devices or care methods to prevent pressure ulcers or to the questions about the organization. The third domain of the instrument did not undergo statistical analysis because the section merely checked the comprehensibility of the questions.

    Phase two. Each of the eight main categories was deemed relevant to pressure ulcer risk. Agreement in the main categories ranged from 94% to 100%. Full agreement (100%) regarding the relevance of the main categories was recorded in mobility, urinary incontinence, and sensory perception. The lowest agreement figure (94%) was recorded for mental status and skin condition (see Table 3).

    High agreement scores were recorded for question clarity and nurses’ ability to define the subcategories (n = 32) (see Table 3). However, the agreement score for clarity of subcategories was lower in almost every item compared to the nurses’ ability to define the subcategories. Agreement regarding nurse ability to define subcategories was less than clarity of subcategories in only four subcategories.

    Incomplete responses were frequent in the expert analysis of the device items. For each question concerning second question domain relevance or clarity, at least seven (14%) responses out of 50 were missing. The items concerning question clarity dealing with down mattress pad, gel-filled sitting cushion, or water-filled sitting cushion were most frequently (44%) incomplete (see Table 4). In almost all questions, the respondents demonstrated more than 70% agreement with both relevance and clarity of the items. Agreement levels greater than 90% were reached in 10 questions concerning relevance and in one concerning clarity (hydraulic hoisting device). Full agreement (100%) was reached in the relevance of air mattresses. The highest agreement scores were recorded in auxiliary devices (technical devices) and methods. Questions where the level of agreement fell below 70% concerned the clarity (61%) or relevance (63%) of questions regarding massaging skin areas susceptible to pressure and the relevance of using ordinary hospital mattresses (67%) or antithrombotic agents (68%). Agreement for relevance (mean 87%) was higher than for clarity (mean 82%).

Discussion

    This article describes the process used to develop an instrument for pressure ulcer assessment in long-term care. Earlier risk assessment scales did not take into account staff number, staff structure, or preventive devices. The instrument under discussion includes these elements. The importance of these elements needs to be addressed further by, for example, conducting another study with a larger number of patients, nurses, and organizations. The purpose of the present study was to discover whether the elements in the assessment tool were relevant in describing the risk of pressure ulcers in the long-term care environment. Therefore, patient information was not analyzed; only questions that solicited this information were assessed.

    The first version of the instrument was developed by combining the contents of 21 earlier scales and incorporating additional risk factors found in the literature. The content validity of the items selected for the instrument were pilot-tested in two phases to ensure the instrument would yield the information desired and that questions represented patient risk factors and preventive devices in use. Both phases also examined the structure of the instrument, assessing whether all dimensions and factors relevant to the phenomena surveyed were addressed.

    Pilot testing demonstrated that the present findings support the main categories defined in earlier risk assessment scales (eg, activity, nutrition, urinary incontinence, and sensory perception). In previously developed instruments, incontinence,^21,37-40 physical condition,³⁷ shear forces,^39,40 neurological status,^40,41 nutritional status, activity, mobility,³⁸ skin type,⁴² and mental status³⁸ were found to correlate strongly with pressure ulcer risk. An additional section of the new instrument addressed indications that organizational factors may influence the occurrence of pressure ulcers.

    The new instrument pilot test included questions regarding staff number and structure mainly to document respondent understanding. The first phase produced information about the use of the instrument in practice. Although only registered nurses were included to ensure the sample represents a homogenous educational background, complete agreement regarding main categories items was not achieved. In one ward, urinary incontinence presented with the lowest agreement percentage in this study, even though among the other wards it was the main category that most often achieved 100% agreement. This result is probably due to the fact that the nurses completed the forms on different days when the patients’ incontinent status could have changed. Before the second phase of pilot testing, the instrument was revised; the main category “appetite” was excluded (the main category “nutrition” remained) and clarifying examples were added to some subcategories.

    Nurses pilot-testing the instrument in phase two can be regarded as a reliable sample in estimating the relevance of the main categories, the clarity of the subcategories, devices and care methods, and nurses’ ability to define subcategories — almost half of them were in charge of wound healing in their respective units. Most of them indicated they treated pressure ulcer patients daily or weekly and felt they were able to assess patient risks for developing pressure ulcers at least “very well.” The respondents’ mean age and the average time served in professional nursing, noted in Methods, were at a slightly lower level than that of the Finnish tissue viability nurse (mean age 43 years, average working years in healthcare 16.5 years).⁴³

    The expert analysis of phase two demanded more of the respondents than the first phase because only the former surveyed the level of difficulty of the survey questions. From the answers, it was obvious that the respondents considered the form “ready-to-go” as an instrument for pressure ulcer assessment. This conclusion seems justified because the respondent criticism, documented under “general comments” at the end of the questionnaire, merely reflected the desire for an easier-to-complete instrument. The patient risk question domain was found easier to complete than the device and method areas, which may explain the missing data/unanswered questions. High numbers of missing data in the devices and methods areas also might be explained by the lack of corresponding devices or methods available to the nurses in their facility. Further, the questions may have seemed repetitive because the responses concerning mattresses, mattress pads, and cushions were similar.

    The instrument was revised according to the feedback received from the experts’ analysis. The technical devices and methods sections remained mainly unchanged with the exception of the corrections made to avoid ambiguity regarding the two technical devices. Also, “real sheepskin” was deleted in the devices section. The items concerning bed, mattresses, mattress pads, and sitting cushions were revised into open-ended questions. The third section of the instrument was kept in its original form (phase one) because head nurses clearly understood those questions and answers could be documented from the hospital’s statistics.

Limitations

    The phases of the study appeared to serve as an appropriate way to collect information quickly and provided a suitable method to examine how the risk assessment instrument functions in practice. The reliability and the validity of the instrument were improved as a result of the pilot test but further study is necessary to document whether desired outcomes (ie, is risk being accurately predicted?) are achieved. In this study, the statistical validity tests could not be carried out because the number of cases was too low. However, the instrument is not yet in its final form. Further data collection with a large number of patients in long-term care should be performed to enhance the caliber of the test.

Conclusion

    Pressure ulcer risk assessment is an important part of the overall healthcare regimen. To address the need for a more complete assessment instrument, Finnish researchers designed and tested a more comprehensive assessment tool than those previously in use. The new instrument incorporates tested instruments and information from more recent research and is based on the fact that the scores exceeding a certain threshold within an area of patient risk activate other areas of the instrument. Most importantly, the new risk assessment tool can help maintain clinician awareness of the many facets of pressure ulcer prevention and management to the betterment of patient care.

1. Collier M. Blanching and non-blanching hyperaemia. A review of terms and definitions involved in the identification of risk of pressure ulcer development. J Wound Care. 1999;8(2):63–64.

2. Cullum N, Deeks J, Fletcher A, Sheldon T, Song F. Preventing and treating pressure sores. Quality in Health Care. 1995;4:289–297.

3. Ek A-C, Boman G. A descriptive study of pressure sores: the prevalence of pressure sores and the characteristics of patients. J Adv Nurs. 1982;7:51–57.

4. Rodriguez G, Murphy K. Current trends in pressure ulcer research. Crit Rev Phys Rehabil Med. 1996;8(1):1–18.

5. O’Sullivan K, Engrav L, Maier R, Pilcher S, Isik F, Copass M. Pressure sores in acute trauma patient: incidence and causes. J Trauma. 1997;42(2):275–278.

6. Lepistö M, Eriksson E, Hietanen H, Asko-Seljavaara S. Patients with pressure ulcer in Finnish hospitals. Int J Nurs Pract. 2001;7(4):280–287.

7. Clark M. Pressure ulcer prevention. In: Morison M, van Rijswijk L (eds). The Prevention and Treatment of Pressure Ulcers. Edinburgh,UK: Mosby;2001:75–78.

8. Dealy C. The Care of Wounds. A Guide for Nurses. London, UK: Blackwell Science;1999.

9. Agency for Health Policy and Research. Clinical Practice Guideline Number 3: Pressure ulcers in adults. Publication No 92-0047. Rockville, Md; Agency for Health Policy and Research;1992.

10. Wiechula R. Pressure sores — part II: management of pressure related tissue damage. Best Practice. Evidence Based Practice Information Sheets for Health Professionals.1997;1:1–6.

11. Klitzman B, Kalinowski C, Glasofer S, Rugani L. Pressure ulcers and pressure relief surfaces. In: Granick M, Long C, Ramasastry S (eds). Clin Plast Surg. 1998;25(3):443–450.

12. Langemo D, Olson B, Hunter S, Hanson D, Burd C, Cathcart-Silverberg T. Incidence and prediction of pressure ulcer in five patient care settings. Decubitus. 1991;4:25–36.

13. Javitz H, Ward M, Marrens L. Major costs associated with pressure sores. J Wound Care. 1998;7:286–290.

14. Stakes. Care Register (HILMO). Helsinki: Stakes (National Research and Development Centre for Welfare and Health);2000. Available at: www.stakes.fi/english/index.html.

15. Schoonhoven L, Haalboom JR, Bousema MT, et al. Prospective cohort study of routine use of risk assessment scales for prediction of pressure ulcers. BMJ. 2000;325(7368):797.

16. Shannon M. Pressure sores. In: Norris C (ed). Concept Clarification in Nursing. Rockville, Md: Aspen;1982:357–382.

17. Clark M, Farrar S. Comparison of pressure sore risk calculators. In: Harding K, Leaper D, Turner T (eds). Proceedings of the 1st European Conference on Advances in Wound Management. Cardiff, 4-6.9.1991 London, UK: Macmillan

18. Birtwistle J. Pressure sore formation and risk assessment in intensive care. Care of the Critically Ill. 1994;10:154–159.

19. Lowery M. A pressure sore risk calculator for intensive care patients: “the Sunferland experience.” Intensive Crit Care Nurs. 1995;11(6):344–353.

20. McCormack H. A pressure sore risk scale for use with older people. Prof Nurse. 1996;11:673–676.

21. Norton D, McLaren R, Exton-Smith A. An Investigation of Geriatric Nursing Problems in Hospital. Edinburgh, UK: Churchill Livingstone;1975.

22. Gosnell D. An assessment tool to identify pressure sores. Nurs Res. 1973;22(1):55–59.

23. Abruzzese R. Early assessment and prevention of pressure sores. In: Lee B, ed. Chronic Ulcers of The Skin. New York, NY: McGraw-Hill Book Company;1985.

24. Braden B, Bergstrom N. Clinical utility of Braden scale for predicting pressure sore risk. Decubitus. 1989;2(3):44–51.

25. Waterlow J. Pressure sores: a risk assessment card. Nurs Times. 1985;81(48):49–55.

26. Andersen K, Jensen O, Kvorning S, Bach E. Prevention of pressure sores by identifying patients at risk. BMJ. 1982;284:1370–1371.

27. Williams C. A comparative study of pressure sore prevention scores. J Tissue Viability. 1992;2(2):64–66.

28. Salzberg C, Byrne D, Cayten C, van Niewerburgh P, Murphy J, Viehbeck M. A new pressure ulcer risk assessment scale for individuals with spinal cord injury. Am J Phys Med Rehabil. 1996;75(2):96–104.

29. Stotts N. Age-specific characteristics of patients who develop pressure ulcers in the tertiary-care setting. Nurs Clin North Am. 1987;22(2):391–398.

30. Barratt E. Putting risk calculators in their place. Nurs Times. 1987;83(6):65–70.

31. Lepistö M. Pressure ulcer risk assessment in long-term care. Developing an instrument. Doctoral thesis. Department of Nursing Science. Annales Universitatis Turkuensis D 588. University of Turku. Painosalama Oy. Turku.

32. Eriksson E, Hietanen H, Asko-Seljavaara S. Prevalence and characteristics of pressure ulcers: a one-day population in Finnish hospitals. Clin Nurse Spec. 2000;14(3):119–125.

33. Lepistö M, Eriksson E, Hietanen H, Asko-Seljavaara S. Prevention of pressure ulcers in acute and long-term care facilities in Finland: results of a survey. Ostomy Wound Manage. 2000;46(6):30–41.

34. Gosnell D. Assessment and evaluation of pressure sores. Nurs Clin North Am. 1987;22(2):399–416.

35. Ek A-C. Prevention, treatment and healing of pressure sores in long term care patients. Scandinavian Journal of Caring Science. 1987a;1:7–13.

36. Ek A-C. Prediction of pressure sore development. Scandinavian J Caring Sci. 1987b;1:77–84.

37. Goldstone L, Goldstone J. The Norton score: an early warning of pressure sores ? J Adv Nurs. 1982;7:419–426.

38. Pajk M, Craven G, Cameron-Barry J, Shipps T, Bennum N. Investigating the problem of pressure sores. J Gerontol Nurs. 1986;12(7):11–16.

39. Tourtual D, Riesenberg L, Korutz C, et al. Predictors of hospital acquired heel pressure ulcer. Ostomy Wound Manage. 1997;43(9):24–40.

40. Haalbom JR, den Boer J, Buskens E. Risk-assessment tools in the prevention of pressure ulcers. Ostomy Wound Manage. 1999;45(2):20–26,28,30–34.

41. Carlson E, Kemp M, Shott S. Predicting the risk of pressure ulcers in critically ill patients. Am J Crit Care. 1999;8(4):262–269.

42. Pang S, Wong T. Predicting pressure sore risk with the Norton, Braden and Waterlow scales in a Hong Kong rehabilitation hospital. Nurs Res. 1998;47(3):147–153.

43. Seppänen S, Iivanainen A. Haavahoitaja 2000 (The tissue viability nurse 2000). Oulainen:Painoykkönen Ky;2000.