Skin Breakdowns in Acute Care Pediatrics (OL]DEHWK�&��6XGGDE\��6FRWW�%DUQHWW��/RUQD�)DFWHDX��Pediatr Nurs. 2005; 31 (2): 132-138.

$EVWUDFW�DQG�,QWURGXFWLRQ�$EVWUDFW�3XUSRVH� To develop a simple, single-page measurement tool that evaluates risk of skin breakdown in the pediatric population and apply it to the acutely hospitalized child. 0HWKRGV� Data were collected over a 15-month period from 347 patients on four in-patient units (PICU, medical-surgical, oncology, and adolescents) on skin breakdown using the AHCPR staging guidelines and compared to the total score on the Starkid SkinScale in order to determine its ability to predict skin breakdown. )LQGLQJV� The inter-rater reliability of the Starkid Skin Scale was r2 = 0.85 with an internal reliablity of 0.71. The sensitivity of the total score was low (17.5%) but highly specific (98.5%). The prevalence of skin breakdown in the acutely hospitalized child was 23%, the majority (77.5%) occurring as erythema of the skin. Buttocks, perineum, and occiput were the most common locations of breakdown. Occiput breakdown was more common in critically ill (PICU) patients while diaper dermatitis was more common in the general medical-surgical population. &RQFOXVLRQV� The Starkid Skin Scale may be useful to pediatric nurses in assessing which patients require intensive prevention measures. Further research is needed as to which of these specific techniques are effective in the prevention of skin breakdown in children.

,QWURGXFWLRQ�Excellent skin care is a hallmark of quality nursing care. Prevalence of skin breakdown has become a standard by which hospitals evaluate themselves, and the American Nurses Association (ANA) identifies it as one of the unique nurse-sensitive outcome measures (ANA, 1995). There is a considerable body of literature on skin breakdown in the adult population including national clinical practice guidelines on pressure ulcers (Bergstrom et al.,1994), national prevalence studies (Amlung, Miller, & Bosley, 2001), and well-respected, valid tools to predict pressure ulcers (Bergstrom, Braden, Laguzza, & Holman, 1987). The Wound, Ostomy and Continence Nurses Society has promoted the consistent description (staging) of skin breakdown (Bergstrom et al., 1994) (see Table 1 ). Nationally accepted measurement tools and clinical practice guidelines have also been developed for the neonatal population (Lund et al., 2001). Yet the key factors that contribute to skin breakdown in children differ from those for the premature infant or adult populations making development of a risk assessment tool specifically designed for children vital to determining the quality of pediatric skin care. 7DEOH����6WDJLQJ�RI�6NLQ�%UHDNGRZQ�

�

5HYLHZ�RI�WKH�/LWHUDWXUH�Okamoto and colleagues (1983) report of decubiti, lacerations, burns, abrasions, and dermatitis in 43% of children with myelomeningocoele observed for skin breakdown annually (for 20 years) offers clear evidence that skin breakdown is a pediatric issue. Okamoto found that the causes of skin breakdown were pressure (42%), friction from casts or orthotic devices (23%), moisture on the skin from soiling (23%), and excessive activity (10%). Children with mental retardation, large head size, kyphoscoliosis, and chronic soiling showed higher levels of breakdown (Okamoto, Lamers, & Shurtleff, 1983). Escher Neidig’s research on pressure ulcers in children after open-heart surgery found that children have areas of breakdown that differ from the classic adult pattern of coccyx, sacrum, and heels. Seventeen percent of children had pressure ulcers, with 10 occurring on the occiput and one on the heel (Escher Neidig, Kleiber, & Oppliger, 1989). Solis and colleagues (1988), using 13 healthy children (aged 10 weeks to 13 years) lying on a pressure sensitive mattress, also found the highest pressure under the occipital area at 59 mmHg. Gershan and Esterly (1993) reported scarring alopecia on the occiput in a case review of five neonates after ECMO. Vasopressor therapy and diminished tissue perfusion were common in all the cases. More recently, Zollo et al. (1996) reported the results of a prospective, matched, controlled study of skin breakdown among 271 admissions to a 14-bed pediatric intensive care unit (PICU). Of those studied, 26% had breakdown at locations including the nose (28%), buttocks/groin (14%), occiput (12%), back (7%), pulse ox site (6%), chest (6%), face (6%) neck (5%), ear (4%), and heels (3%). Although the researchers analyzed many variables, the only predictors of skin breakdown were white race (odds ration [OR] 1.13) and PRISM score (indicator of risk of mortality) on admission (OR 3.95). Waterlow (1997) presented data on 302 patients admitted to five British hospitals where 17 children (5.6% of admissions) developed 33 pressure sores. There was a mix of orthopedic, medical-surgical, rehabilitation, and intensive care patients from neonate to 16 years of age. The most common areas of breakdown were heels (15%) and leg, elbow, buttocks and ankle (12% each). In a retrospective review, Pallija, Mondozzi, and Webb (1999) found that over a 4- year period, 994 hospital days were used to care for skin breakdown in children with spina bifida, which underscores the importance of assessment for pressure ulcers in specific populations. Samaniengo (2003) retrospectively analyzed the pressure ulcers in 50 children followed in a wound care clinic. Breakdown occurred on feet (27%), lower leg (26%), heels (11%), and upper legs (8%) in a population of patients with myleogysplasia. The risk factors identified from the documentation were paralysis, insensate areas, high activity, and immobility. Many of the patients had orthotic or prosthetic devices, or plaster casts. Others were confined to wheelchairs. Wheelchair use increased the risk of a friction injury from areas of rubbing or as the child outgrew the chair. Curley, Quigley, and Lin (2003) reported on a prospective multi-center study of pressure ulcers found in PICU patients between the ages of 3 weeks and 8 years at three study sites with a total of 322 patients. Incidence of pressure ulcers (of any stage) was 27%. The incidence by stage was 70% stage I, 27% stage II, and 3% stage III. The location of the stage II or III ulcers involved the head (occiput and ears) (32%) and the sternum, elbow, sacrum, coccyx, and ankles (6.7% each). Factors associated with pressure ulcers included younger age, use of mechanical ventilation, length of mechanical ventilation, use of high frequency oscillatory ventilation, chemical paralysis, vasopressors, total parenteral nutrition, and lower mean arterial pressure. Other than age, all are indicators of the more critically ill child. One finding of particular interest was that patients developed a pressure sore within the first 1-2 days after admission (sores noted on admission were excluded). Excluded from this review were patients with congenital heart disease and any ulcers caused by medical devices (27 additional ulcers) (Curley, Quigley, et al., 2003). A multiple logistic regression model demonstrated that the four factors that most accurately predicted pressure ulcers were mechanical ventilation, Ramsey score (measure of patient’s level of response), mean arterial pressure 50, and Braden Q risk

of skin breakdown score (area under the curve = 0.77) (Curley, Razmus, Roberts, & Wypij, 2003) Despite clear evidence from the literature that skin breakdown is an issue for pediatric patients, there are only two reports of scales to predict risk of skin breakdown in children. Waterlow (1998) described an adaptation of an adult scoring card for children but provided no reliability or validity data. Quigley and Curley (1996) adapted the Braden scale - creating the Braden Q for the pediatric population. Curley et al. (2003) reported that pressure ulcer positive patients have significantly lower mean Braden Q scores ( S < 0.001). Using a cut-off score of 16, the sensitivity of the scale was 0.88 and the specificity 0.58. Although the Braden Q includes seven subscales - mobility, activity, sensory perception, friction/shear, moisture, nutrition and tissue perfusion - only three (sensory perception, mobility, and tissue perfusion/oxygenation) contributed greater than 0.7 to the area under the curve (Curley et al., 2003).

&RQFHSWXDO�)UDPHZRUN�Bergstrom et al. (1987) described a framework of five concepts that contribute either to the intensity of pressure on the skin or to tissue tolerance of that pressure (see Table 2 ). 7DEOH����&RQFHSWXDO�)UDPHZRUN� The pediatric literature validates Bergstrom’s framework. For example, various studies support the concepts of pressure resulting from immobility or decreased sensation (Curley, Quigley, et al., 2003; Escher Neidig et al., 1989; Okamoto et al., 1983; Pallija et al., 1996; Samaniengo, 2003;) and of tissue tolerance, which includes moisture, friction, nutrition and hemodynamic status (Curley, Razmus, et al., 2003; Okamoto et al., 1983; Samaniengo, 2003). Therefore, Bergstrom’s conceptual framework is useful in evaluating skin breakdown in children, and it provides the key concepts we have used for tool development and analysis of breakdown in this study.

0HWKRGRORJ\�'HVLJQ�The work described here began before the publication of the reliability and validity of the Braden Q. We used the Braden Q (Quigley & Curley, 1996) as the basis for developing a simpler, more parsimonious measure of risk of skin breakdown. This tool was designed to serve as a risk assessment guideline that any pediatric nurse could use easily with a minimum

of instruction. It could be used in hospitals or other health care settings to describe quality of care in prevalence studies (Gallagher, 1997) or in daily assessments to identify patients in need of prevention or intervention for skin breakdown. After creating the tool, we collected prevalence of skin breakdown data to determine whether this adaptation of the Braden Q was valid and useful for the pediatric nurse.

'HILQLWLRQ�RI�7HUPV�6NLQ�EUHDNGRZQ�� Changes to intact skin, including nonblanchable erythema without overlying tissue loss, abrasion, mild entry into the skin, and deep and extensive pressure ulcers. 3UHVVXUH�XOFHU�� Damage to the skin resulting in skin loss with a crater (mild or deep) as a result of pressure or friction/shear of skin against other surfaces. 3UHYDOHQFH�� A snapshot of a clinical situation, usually involving an assessment of all patients on one day or at one time. Prevalence is expressed as a percentage of those with "a disease" compared to all patients assessed. It is not equal to incidence because incidence includes a measurement over a period of time.

6HWWLQJ�DQG�6DPSOH�The development of the tool and collection of prevalence data occurred in a single hospital setting of 157 pediatric beds. The hospital is a children’s hospital located within its own facility on the same campus as a large community medical center offering Level I trauma service, inpatient and outpatient surgery center, and substance abuse facilities. The nursing staff ranges from new graduates to nurses with 30 or more years of experience in pediatric nursing. The patients included in data collection were all those present on specific data collection dates in the PICU, pediatric medical-surgical unit, pediatric oncology unit, and adolescent unit. Although the NICU staff collected prevalence data, they did not utilize the same measurement tools and are not reported here. Data were collected for five quarterly skin prevalence surveys resulting in information on 347 pediatric patients. There may have been reports on the same patient in more than one quarter if the child remained in the hospital for more than 3 months. Each incidence was considered a separate assessment of risk. This dataset was then used to evaluate the ability of the measurement tool, The Starkid Skin Scale, to predict risk of skin breakdown by comparing the risk score to actual skin breakdown.

7RRO�'HYHORSPHQW�Two clinical nurse specialists created the measurement tool by rewording and simplifying the concepts presented in the Braden Q (Quigley & Curley, 1996). The goal was to create a one-page, easy-to-use measurement tool that had no areas of overlap and that provided brief, clear descriptors. Past users of the Braden scale had raised questions about the meaning of descriptors for the pediatric population. For example, a baby does not normally walk; thus, would a skin surveyor score a baby lower for a developmental trait? Although the Braden Q addressed pediatric issues, the descriptors are detailed. To simplify the Braden Q further, the categories of mobility and activity were combined because they are closely related in meaning. Finally, bold typeface was added to emphasize the key elements in the scoring of each category (see Table 3 ). A pilot of the Starkid Skin Scale was conducted. No education was provided on the scale because it was designed to be self-explanatory. No changes were made to the scale based on staff nurse views on the clarity of the descriptors. Since prevalence of skin breakdown data was also collected, minor changes were made to the prevalence section of the data form based on comments provided. Four staff nurses, one from each unit of the data collection �

7DEOH����0HDVXUHPHQW�7RRO�������������� hospital (pediatric med-surg, pediatric oncology, adolescents, and PICU), collected all the data. All four nurses had been data collectors for prior prevalence studies that used the Braden adult tool. Each of the nurses was provided with a color guide to the Staging of Skin Breakdown (Bergstrom et al., 1994). All patients were physically assessed for skin breakdown. If breakdown was identified, it was scored using the AHCPR staging scale, location was noted and any interventions described. Each patient was also scored for risk of skin breakdown using the Starkid Skin Scale. In addition, demographic information on age, weight, skin color, admitting service, medical devices attached to the child, skin conditions present upon admission, bed type, and episodes of diarrhea in the prior 24 hours was collected.

�

6WDWLVWLFDO�$QDO\VLV�Continuous data are presented as mean and standard deviation. Categorical data are presented as frequency and percent. Unconditional logistic regression was used to test association of selected parameters with the presence or absence of skin breakdown. All analyses were performed using SAS version 8.2 (Cary, NC). All statistical tests were two-sided with an level set at 0.05. Odds ratios (OR) and 95% confidence intervals (CI) are provided. The OR is the ratio of the odds of disease for a treatment group relative to the odds of disease in a control group. For example, the interpretation of an OR = 1.5 is that the treatment group is at a 50% increased risk (1.5 -1 = 0.5*100 or 50%). An OR is a relative measure in which unity (no difference between groups) is 1. To arrive at a percentage, it is necessary to subtract 1 from the OR and multiply that result by 100. The advantage of using the OR is that any number greater than 1 suggests the treatment or exposed group is at an increased risk of the selected outcome. Additionally, S -values are not necessary to determine statistical significance. Any reported confidence interval that contains 1 is not statistically significant, regardless of any a priori specified Type I (a) error rate. Sensitivity and specificity of the risk assessment for Starkid Skin Scale was determined. Sensitivity refers to the percentage of persons with skin breakdown correctly identified as being at risk for breakdown by the scale (or tool score). Specificity refers to those persons without skin breakdown correctly identified as being at low risk for breakdown.

5HVXOWV�7RRO�'HYHORSPHQW�Inter-rater reliability data was collected on the Starkid Skin Scale using the same four staff nurses and three additional clinical nurse specialists (CNS). Thirty patients were scored by both a staff nurse and a CNS on the same day but at separate times. The results were analyzed using the correlation coefficient between paired observations of total scores ( U 2 = 0.85). The subcategory with the most differences, 10 of the 30 pairs, was nutrition, with an average point difference of 1.4. Internal reliability was good (Cronbach’s alpha: 0.71). Subscales were internally correlated as follows: friction shear, U = 0.65; mobility, U = 0.62; sensory perception, U = 0.56; moisture, U =0.45; tissue perfusion, r=0.30; and nutrition, U = 0.26. Maximal and minimal total scores for the tool on the patients assessed were 24 and 9. Sensitivity and specificity for a maximal score of 24 was 100% and 0%. Sensitivity and specificity for a minimal score of 9 was 0% and 99.6%. Assuming a prevalence of skin breakdown equal to that of our population (23%), any individual with a total score of 15 with low sensitivity (17.5%) but high specificity (98.5%) would have a 74% chance of having a skin breakdown.

5LVN�)DFWRUV�IRU�6NLQ�%UHDNGRZQ�Of the 347 pediatric patients assessed for skin breakdown and scored for features that could identify risk of skin breakdown, 41% were female and 59% male, with 51.3% of patients having olive skin, 34.3% fair skin, and 14.4% black skin. The bed surfaces the patient slept upon included cribs (41%); adult beds (46%); junior beds, which are adult in size but have full side rails the length of the bed (10%); and specialty beds (2%). Additional demographic data are summarized in Table 4 and Table 5 . Table 6 breaks down the patients by admitting service. No skin breakdown on the date of the survey was found in 267 patients, while 80 patients had 100 sites of breakdown (a 23% prevalence of skin breakdown). The majority (77.5%) of the skin breakdown was described as Stage I, redness without a true break in the skin. See Table

7 for level of skin breakdown. The most common areas of breakdown were the buttocks, perineum, and occiput. See Table 8 for areas of breakdown. The most common treatment of skin breakdown was a barrier cream such as Vaseline®, A & DAE, Desitin®, Proshield®, Eucerin®, or other lotions. See Table 9 for all treatment techniques applied. Unconditional logistic regression of demographic features demonstrates that patients with skin breakdown were younger and smaller, with more episodes of diarrhea, more medical devices, and a lower Starkid Skin Scale score (all significant at S 0.05). Patients with skin problems on admission were more likely to have skin breakdown at the time of the assessment. In addition, neurosurgical and cardiac patients had a high frequency of skin breakdown. Patients in the PICU and medical-surgical unit were more likely to have skin breakdown. A further analysis of patients in the PICU showed they had a lower mean Starkid Skin Scale score of 17.5 (median 18, range 9-23). Of the areas of breakdown, 88% of the occipital breakdown and 100% of lesions on toes occurred in PICU patients. Analysis of patients in the medical-surgical unit demonstrated they were younger, with a mean age of 2.8 years (median 1 year), and had the highest rate of diarrhea (mean 1.2 episodes). Of all stages of skin breakdown, the medical-surgical unit had 65% occurring on the buttocks or perineum, with 50% of the stage II lesions occurring on the buttocks or perineum. 7DEOH����'HPRJUDSKLF�)HDWXUHV�RI�3HGLDWULF�3DWLHQWV�$VVHVVHG�IRU�6NLQ�%UHDNGRZQ� 7DEOH����'HPRJUDSKLF�)HDWXUHV�RI�3HGLDWULF�3DWLHQWV�ZLWK�6NLQ�%UHDNGRZQ��E\�3HUFHQWDJH��

7DEOH����$GPLWWLQJ�6HUYLFH�DQG�)UHTXHQF\�RI�%UHDNGRZQ� 7DEOH����/HYHO�RI�%UHDNGRZQ� 7DEOH����$UHDV�RI�%UHDNGRZQ�

7DEOH����7UHDWPHQW�RI�6NLQ�%UHDNGRZQ�����������'LVFXVVLRQ�The Starkid Skin Scale presented here is a reasonable measure of risk of skin breakdown with a high level of specificity (but low sensitivity). This may be useful to pediatric nurses in determining which patients most need preventive measures. Similar to the Braden Q, upon which this new tool was based, the subscales of mobility and sensory perception were important contributors to the tool’s ability to predict breakdown. The subscale of nutrition had the poorest inter-rater reliability and weakest correlation to skin breakdown. This may represent the data collectors’ difficulty with patients who did not perfectly fit the descriptors (i.e. NPO 3 days but not 5 days). So, greater clarity of the descriptors might improve the inter-rater reliability. The contribution of nutrition to skin breakdown may also reflect chronic nutritional deficiencies not found in the acutely hospitalized child. The Starkid Skin Scale did not demonstrate a strong correlation in the area of tissue perfusion, as found by Curley, Quigley and Lin (2003). This may be true due to the small PICU population of this study as compared to the larger sample size in the Braden Q validation study. The findings from the PICU population are consistent with other research in location of breakdown (occiput more than sacrum or coccyx) (Curley, Quigley, et al., 2003; Escher Neidig et al., 1989; Zollo et al. 1996). This study found medical devices were an important risk factor to breakdown. This may explain the higher prevalence of breakdown reported in this PICU at 42% compared with the PICUs in studies by Curley, Quigley et al. (2003) (27%) and Zollo et al., 1996 (26%). The small sample size for the PICU data may skew the prevalence results as well. A strength of our study is that it reports on skin breakdown in the overall hospitalized child population. Other studies, with the exception of Waterlow (1997), do not report on patients in "general" care areas. In addition, previous studies do not cover the significant problem of diaper dermatitis (found here as stage I breakdown in the buttocks and perineum), particularly associated with diarrhea. This may be an obvious finding to the experienced pediatric nurse, but there is no literature on incidence or effective treatments of this type of breakdown

&OLQLFDO�$SSOLFDWLRQV�RI�5HVHDUFK�This study adds to the current knowledge on skin breakdown in the pediatric population, particularly in the acutely hospitalized child, but further research is indicated. The Starkid Skin Scale should be validated in other hospital and non-hospital settings and identification of risk factors should continue. Moreover, the risk factors already identified here and in the literature should be applied to clinical practice guidelines to determine whether specific preventive techniques can change the prevalence rates. Such testing could improve skin care for children. Barrier creams are well known to both nurses and parents but there have been no studies on which ones are most effective or preferred, which would also be invaluable. Skin care is a basic and essential factor in all nursing, but it may not offer a dramatic area for research. Nonetheless, easing the distress of children with skin breakdown and their parents who must care for them offers ample motivation for efforts to find new techniques for solving this key nursing problem. The 3UDFWLFH�$SSOLFDWLRQV�RI�5HVHDUFK section presents reports of research that are clinically focused and discuss the nursing application of the findings. If you are interested in author guidelines and/or assistance, contact Janice S. Hayes, PhD, RN; Section Editor; Pediatric Nursing; East Holly Avenue Box 56; Pitman, NJ 08071-0056; (856) 256-2300 or FAX (856) 256-2345. $FNQRZOHGJHPHQWV Special thanks for their contributions to this work go to Margaret Sourbeer, MSN, RN, Pediatric Intensive Care Clinical Nurse Specialist; Ilona Soboleva, Clinical Nurse II, PICU; Rebecca Nichols, Clinical Nurse IV, Pediatrics Med-Surg; Vivienne Hess, Clinical Nurse II, Pediatric Oncology; Chrisanne Lenzch, Clinical Nurse II, Adolescents; and all of Inova Fairfax Hospital for Children 5HIHUHQFHV�

• American Nurses Association (ANA). (1995) 1XUVLQJ�UHSRUW�FDUG�IRU�DFXWH�FDUH . Washington, DC: ANA.

• Amlung, S.R., Miller, W.L., & Bosley, L.M. (2001). The 1999 national pressure ulcer prevalence survey: A benchmarking approach. $GYDQFHV�LQ�6NLQ��:RXQG�&DUH���� , 297-301.

• Bergstrom, N., Allman, R., Alvarez, O., Bennett, M.A., Carlson, C., Frantz, R., et al. (1994) . Treatment of pressure ulcers: Clinical practice guidelines. Agency for Health Care Policy and Research (AHCPR) (Pub 95-0652). Rockville, MD: U.S. Department of Health and Human Services.

• Bergstrom, N., Braden, B.J., Laguzza, A., & Holman, V. (1987) The Braden scale for predicting pressure sore risk. 1XUVLQJ�5HVHDUFK , �� (4), 206-210.

• Curley, M.A.Q., Razmus, I.S., Roberts, K.E., & Wypij, D. (2003) Predicting pressure ulcer risk in pediatric patients. 1XUVLQJ�5HVHDUFK , 52(1), 22-33.

• Curley, M.A.Q., Quigley, S.M., & Lin, M. (2003) Pressure ulcers in pediatric intensive care: Incidence and associated factors. 3HGLDWULF�&ULWLFDO�&DUH�0HGLFLQH , � (3),284-290.

• Escher Neidig, J.R., Kleiber, C., & Oppliger, R.A. (1989). Risk factors associated with pressure ulcers in the pediatric patient following open-heart surgery. 3URJUHVV�LQ�&DUGLRYDVFXODU�1XUVLQJ��� (4), 99-106.

• Gallagher, S.M. (1997). Outcomes in clinical practice: Pressure ulcer prevalence and incidence studies. 2VWRP\�:RXQG�0DQDJHPHQW , �� (1), 28-40.

• Gershan, L.A., & Esterly, N.B. (1993). Scarring alopecia in neonates as a consequence of hypoxaemia-hypoperfusion. $UFKLYHV�LQ�'LVHDVH�LQ�&KLOGUHQ����� 591-593.

• Lund, C.H., Kuller, J., Lane, A.T., Lott, J.W., Raines, D.A., & Thomas, K.K. (2001). Neonatal skin care: Evaluation of the AWHONN/NANN research-based practice project

on knowledge and skin care practices. -RXUQDO�RI�2EVWHWULF��*\QHFRORJLF�DQG�1HRQDWDO�1XUVLQJ���� (1), 30-51.

• Okamoto, G.N., Lamers, J.V., & Shurtleff, D.B. (1983). Skin breakdown in patients with myelomeningocele. $UFKLYHV�LQ�3K\VLFDO�0HGLFLQH�DQG�5HKDELOLWDWLRQ����� 20-23.

• Pallija, G., Mondozzi, M., & Webb, A.A. (1999). Skin care of the pediatric patient. -RXUQDO�RI�3HGLDWULF�1XUVLQJ���� (2), 80-87.

• Quigley, S.M., & Curley, M.A.Q. (1996). Skin integrity in the pediatric population: Preventing and managing pressure ulcers. -RXUQDO�RI�WKH�6RFLHW\�RI�3HGLDWULF�1XUVLQJ��� (1), 7-18.

• Samaniengo, I.A. (2003). A sore spot in pediatrics: Risk factors for pressure ulcers. 3HGLDWULF�1XUVLQJ , �� (4), 278-282.

• Solis, I., Krouskop, T., Trainer, N., & Marburger, R. (1988). Supine interface pressure in children. $UFKLYHV�RI�3K\VLFDO�0HGLFLQH�DQG�5HKDELOLWDWLRQ���� , 524-526.

• Waterlow, J. (1998). Pressure sores in children: Risk assessment. 3DHGLDWULF�1XUVLQJ , � (6), 21-24.

• Zollo, M.B., Gostisha, M.L., Berens, R.J., Schmidt, J.E., & Weigle, C.G.M. (1996). Altered skin integrity in children admitted to a pediatric intensive care unit. -RXUQDO�RI�1XUVLQJ�&DUH�4XDOLW\ , �� (2):62-67.