- Email: [email protected]
Become the PPUPET Master: Mastering Pressure Ulcer Risk Assessment With the Pediatric Pressure Ulcer Prediction and Evaluation Tool (PPUPET)
Journal of Pediatric Nursing (2015) 30, 598–610
Become the PPUPET Master: Mastering Pressure Ulcer Risk Assessment With the Pediatric Pressure Ulcer Prediction and Evaluation Tool (PPUPET)
David J. Sterken MN, CPNP a,⁎, JoAnn Mooney BSN, RN CPN b , Diana Ropele MSN, RN, CCRN c , Alysha Kett BS, MS d , Karen J. Vander Laan PhD, MSN, RN e a
Pediatric Services, Helen DeVos Children's Hospital, Grand Rapids, MI Quality Improvement, Helen DeVos Children's Hospital, Grand Rapids, MI c Pediatric Trauma Services, Helen DeVos Children's Hospital, Grand Rapids, MI d Spectrum Health, Grand Rapids, MI e Spectrum Health Hospitals, Center for Nursing Practice & Development, Grand Rapids, MI b
Received 24 February 2014; revised 2 October 2014; accepted 7 October 2014
Key words: Hospital acquired pressure ulcer (HAPU); Pressure ulcer; Pediatric pressure ulcers; PPUPET
Hospital acquired pressure ulcers (HAPU) are serious, debilitating, and preventable complications in all inpatient populations. Despite evidence of the development of pressure ulcers in the pediatric population, minimal research has been done. Based on observations gathered during quarterly HAPU audits, bedside nursing staff recognized trends in pressure ulcer locations that were not captured using current pressure ulcer risk assessment tools. Together, bedside nurses and nursing leadership created and conducted multiple research studies to investigate the validity and reliability of the Pediatric Pressure Ulcer Prediction and Evaluation Tool (PPUPET). © 2015 Elsevier Inc. All rights reserved.
NOSOCOMIAL PRESSURE ULCERS are a serious, debilitating, and preventable complication in all inpatient populations. Similar to adult patients, acutely ill infants and children are at risk for pressure ulcers (Bernabe, 2012; Kottner, Wilborn, & Dassen, 2010; Parnham, 2012; Quigley & Curley, 1996). However, because of the anatomical and physiological differences in infants and children, the most common sites for pressure ulcer development are different than the adult population. For example, the head makes up a greater proportion of the total body weight and surface; thus, the occipital region of the scalp is the most common site of ulceration for infants and children (Willock, Harris, Harrison, & Poole, 2005). Infants and children are also prone to develop pressure ulcers on other bony prominences as well as any area where external medical devices (such as ⁎ Corresponding author: David J. Sterken, MN, CPNP. E-mail address: [email protected]. http://dx.doi.org/10.1016/j.pedn.2014.10.004 0882-5963/© 2015 Elsevier Inc. All rights reserved.
pulse oximetry probes, oxygen tubing, intravenous hubs, braces or other tubes) are placed (Kottner et al., 2010; Waterlow, 1997; Willock, Baharestani, & Anthony, 2009). Some conditions that place children at greater risk for pressure ulcer development include extracorporeal membrane oxygenation (ECMO), paraplegia, myelomeningocele, large head size, kyphosis, developmental delay, and chronic fecal and urinary soiling (Noonan, Quigley, & Curley, 2006). Despite evidence of the development of pressure ulcers in the pediatric population, relatively minimal research has been done to develop pediatric pressure ulcer risk assessment tools. The purpose of this article is to introduce and report testing of the Pediatric Pressure Ulcer Prediction and Evaluation Tool (PPUPET), a pediatric-focused skin risk assessment instrument developed by bedside nurse leaders in response to finding pressure ulcers in our patient population. The use of standardized tools in nursing promotes consistency among caregivers and facilitates communication
Become the PPUPET Master regarding evaluation and management of clinical decisions and treatments for various clinical conditions. In pediatrics, however, standardized tools are often modifications of adult tools and do not always address the unique considerations of the pediatric patient. Standardized tools must also be re-evaluated and revised periodically as healthcare changes and the language used to describe clinical conditions becomes more uniform. Standardized tools must be easy to use and intuitive for the bedside nurse; otherwise, the tool can add burden which takes away from patient care. Evaluation testing of standardized tools is also important to ensure that the tool consistently measures the clinical condition and that appropriate assessment criteria are being considered. Inter-rater testing, which assesses the degree to which different raters independently concur in their observations of what is being measured, is used to report a tool's reliability (Polit & Beck, 2013). Validity—the degree to which an instrument measures what it is intended to measure—can be established by comparing the items in a tool to another reliable, valid instrument measuring similar concepts (Polit & Beck, 2013). When a tool is used in clinical practice at a time when outcome conditions can also be assessed, calculations of sensitivity—the tool's ability to identify patients who have a true positive condition—and specificity—the tool's ability to identify patients who have a true negative condition—help clinicians appraise the usefulness of the standardized tool (MedCalc Software, 2014). For the PPUPET, sensitivity measures the proportion of children rated at risk who actually have a pressure ulcer, while specificity measures the proportion of children rated not at risk who actually have no pressure ulcers. Lastly, positive and negative predictive value can also be calculated for an instrument (MedCalc Software, 2014). For the PPUPET, the positive predictive value measures the probability that the risk for pressure ulcer development is present when the child is assessed as being at risk for pressure ulcer development. Negative predictive value measures the probability that the risk for pressure ulcer development is not present when the child is assessed as not being at risk.
Purpose and Background The purpose of this article is to report the results of recent retrospective and prospective studies of the PPUPET which have evaluated its usefulness as a risk assessment tool for the development of pressure ulcers in the pediatric population. As background, we will describe the development and initial reliability/validity testing of the PPUPET and then discuss the more recent studies.
Bringing the PPUPET to Life In the fall of 2004, members of the children's hospital nursing leadership team began a conversation regarding the prevalence of pressure ulcers in our pediatric patient population. Our adult hospital colleagues already had been performing hospital acquired pressure ulcer (HAPU) audits to develop strategies for the prevention and treatment of
599 pressure ulcers. In order to establish a baseline for prevalence of pressure ulcers, our first skin audit was completed in November 2004. A complete head to toe skin assessment was conducted on 49 hospitalized pediatric patients on a single day. Results of our audit demonstrated a prevalence of pressure ulcers in 9% of these patients. In May 2005, the next skin audit was performed on 59 patients. Results of the audit demonstrated that 15% of the patients had a pressure ulcer. At this time the nursing leadership team realized that pressure ulcers were indeed a problem at the children's hospital and began to formulate a strategy to address the issue. Nursing leadership recruited bedside nurses from each clinical unit to form a team to address the prevalence of pressure ulcers in our pediatric patient population. Members of the “Skin Team” were assigned to conduct biannual skin audits. Following the January 2006 skin audit of 74 pediatric patients—which yielded an alarming 11% prevalence of stage 1 (n = 8) and 1% stage 2 (n = 1) pressure ulcers1 and several patients having N 1 pressure ulcer—the skin team identified the following concerns: • No standardized tool was being used to identify patients at risk for pressure ulcer development. • No pediatric plan of care had been developed for the prevention and management of pressure ulcers in children. • The nursing Kardex2 did not reflect nursing orders for turning and basic skin care. • Nurses were unaware that they could initiate nursing orders/interventions to prevent the development of pressure ulcers. • Documentation regarding nursing interventions (e.g., turning and positioning) for the prevention of pressure ulcers was lacking or inconsistent. • Nurses were unaware of where or how to document positioning in new electronic nursing documentation. • Care related to devices (e.g., repositioning pulse oximetry probes) was inconsistently documented.
Based on the identified concerns, the team decided to review standardized pressure ulcer risk assessment tools and develop a standard of practice for assessment and intervention to prevent pediatric pressure ulcers. Starting in March 2006, skin team members began reviewing the existing evidence. We first reviewed the Braden Scale for Predicting Pressure Sore Risk (Bergstrom, Braden, Laguzza, & Holman, 1987) [Braden], which was the tool used by our 1
Pressure ulcer staging is based on the National Pressure Ulcer Advisory Panel's (NPUAP) definitions of stages 1 through 4. Stage 1 is the least severe (non-blanchable erythema), and stage 2 is a partial-thickness ulcer. The current version of the Quick Reference Guide (European Pressure Ulcer Advisory Panel and National Pressure Ulcer Advisory Panel, 2009), now includes stages 1 to 4 and deep tissue injury and unstageable categories. 2 The paper Kardex was a primary tool utilized by bedside nursing staff to communicate pertinent patient-specific information, direct nursing care based on physician orders, facilitate nurse-to-nurse communication, and track patient significant events (VanderKooi, Blackport, & Vander Laan, 2011).
600 colleagues caring for adult patients. The scale has 6 subscales: sensory perception, moisture, activity, mobility, nutrition, and friction & shear. Each subscale is mutually exclusive. Each item of the subscale has a minimum score of 1 (more risk) and a maximum score of 3 or 4 (less risk). The range of scores is 6–23. In the acutely ill adult population cut-off points of 16 to 18 are used to identify patients at risk for developing pressure ulcers. At a score of 16 the Braden Scale has a sensitivity of 0.83 and specificity of 0.64 in predicting pressure ulcer formation (Bergstrom, Demuth, & Braden, 1987). Next we looked at the Braden Q Scale for Predicting Pressure Ulcer Risk (Quigley & Curley, 1996) [Braden Q], which was validated for use in pediatric patients from 3 weeks to 8 years of age (Curley, Razmus, Roberts, & Wypij, 2003). They made changes to the original scale related to developmental needs and treatment interventions common to the pediatric patient and added a seventh subscale for tissue perfusion and oxygenation. The minimal score for each subscale is 1 (more risk), and the maximum score is 4 (less risk). Potential scores range from 7 to 28 points, the lower the score the higher the risk for ulcer development. Content validity was established by a group of pediatric nurses with special interest in skin issues. Confidence intervals indicated that children scoring b 23 were at moderate or high risk for pressure ulcers (Quigley & Curley, 1996). An initial study to establish the predictive validity of the tool revealed that at a score of 16 the sensitivity was 0.88 and the specificity was 0.58 (Curley et al., 2003). Concepts from these validated tools were recognized as important to pediatric pressure ulcer risk assessment. Members of the committee felt that the Braden and Braden Q did not identify some key risk factors unique to our population and confirmed by our skin audits. First, a compromised skin condition was not identified as a risk factor for pressure ulcer development. Second, external medical devices—such as pulse oximetry probes, intravenous hubs/protective devices, and orthopedic equipment— were not recognized as significant causes of pressure ulcers. In addition, use of the Braden and Braden Q required two different tools based on the age of the patient. Lastly, nursing staff indicated that the Braden and Braden Q tools were not intuitive in scoring, as the higher the score the lower the risk of developing a pressure ulcer. We then explored the possibility of adding additional subscales to the Braden Q in a conversation with Dr. Curley (personal communication, October 2006). However, at that time, modification of the Braden Q was not an option due to copyright restrictions. Subsequently, a decision was made to develop a tool specific for our population, using valid and reliable concepts related to pressure ulcer risk from the literature, our institution's sources of pressure ulcers, and recommendations from clinical nursing staff. The Pediatric Pressure Ulcer Prediction and Evaluation Tool (PPUPET) was created in 2007 (Figure 1) for use with children from birth to 18 years, with the exception of premature infants cared for in the neonatal intensive care unit. The PPUPET
D.J. Sterken et al. covers 9 subscales that are mutually exclusive. The subscales include mobility, activity, sensory perception, moisture, external medical devices, friction/shear, tissue perfusion & oxygenation, skin condition and nutrition. Each item of the subscale has a minimum score of 1 (less risk) and a maximum score of 2 or 3 (more risk). The range of scores is 9–26. A score of 18–26 indicates that the patient is at risk for developing a pressure ulcer. In addition to the total score, a score of 3 in any subscale or a 2 in the nutrition subscale also puts a patient at risk. All children at risk of pressure ulcer development require implementation of a plan of care for the prevention of skin breakdown. The PPUPET included external medical devices and skin condition as two subscales not previously identified in the literature. To make the tool intuitive, higher scores reflected greater risk of pressure ulcer development. The tool was piloted and adjusted to incorporate the clinical nursing staff's suggestions regarding reference text for each subscale. In conjunction with the PPUPET, a plan of care (Figure 2) was created to identify specific key interventions to be implemented for risks identified by high subscale scores. Education was then provided to all staff regarding the risks of development of pressure ulcers, especially those related to medical devices; interventions for prevention of pressure ulcers, including repositioning of medical devices if possible; and instructions for documentation of interventions,
Testing the PPUPET's Performance Next the skin team, led by a clinical nurse specialist, designed a validation study to compare the PPUPET to current validated scales identified in the literature and to document inter-rater reliability. For our study we used the Braden for patients age 5 years and older and the Braden Q for patients under 5 years of age, as provided in the Institute for Healthcare Improvement's How-to-guide Pediatric Supplement Preventing Pressure Ulcers (Pediatric Affinity Group, last updated 02-01-2014). In November 2008, the pressure ulcer risk prediction from the items assessed in the PPUPET was compared to the Braden or Braden Q scales during that quarter's HAPU prevalence audit of 48 pediatric patients. The PPUPET identified 7 patients (14.7%) as high risk for pressure ulcer development; 2 of these patients (4.2%) were found to actually have pressure ulcers. This computes to a sensitivity of 100.00% (95% CI 19.29, 100.00), specificity of 89.13% (95% CI 76.48, 96.33), a positive predictive value of 28.51% (95% CI 4.52, 70.73), and a negative predictive value of 100.00% (95% CI 91.31, 100.00). The percentage of agreement between the PPUPET and the Braden Q (for patients b 5 years of age) was 97% (n = 30) with a Kappa3 of 3 The Kappa statistic corrects for the number of times agreement might be expected to happen by chance, so it is a more accurate reflection of the actual (true) agreement (Sim & Wright, 2005).
Become the PPUPET Master
Figure 1
601
Pediatric pressure ulcer prediction and evaluation tool (PPUPET).
0.650. The percentage of agreement between the PPUPET and the Braden (for patients ≥ 5 years of age) was 83% (n = 18) with a Kappa of 0.557. Scoring of patients as high risk on the PPUPET was significantly related to scoring as high risk on the Braden Q (χ2 (1) = 14.483, p b 0.001) and the Braden (χ2 (1) = 5.716, p = 0.017). Inter-rater reliability was tested by comparing staff nurses' most recent PPUPET scores to Skin Champion's PPUPET scores on the day of the hospital skin audit. There was a significant difference (t (43) = − 2.922, p = 0.005) between the staff nurses' PPUPET score (mean = 12.78,
std dev = 3.649) and the Skin Champions' PPUPET score (mean = 13.49, std dev = 3.764), which led to further staff education and revisions of the reference text. The current version of the PPUPET was adopted as a standard nursing assessment in 2009. Studies to Validate the PPUPET In 2012, two studies—one retrospective and one prospective—were designed to validate the PPUPET. The purpose of the retrospective study was to review the medical
602
Prevention of skin breakdown interdisciplinary plan of care (Page 1 of 4).
D.J. Sterken et al.
Figure 2
603
Figure 2 (continued).
Become the PPUPET Master
D.J. Sterken et al.
Figure 2 (continued).
604
605
Figure 2 (continued).
Become the PPUPET Master
606 records of pediatric patients hospitalized between January 1, 2010 to December 31, 2012 who had been assessed using the PPUPET to evaluate the association between PPUPET score and pressure ulcer development. The purpose of the prospective study was to compare the predictive validity of the PPUPET to the predictive validity of the Braden or Braden Q and to a new tool, the Glamorgan (Willock, Anthony, & Richardson, 2008).
Methodology Setting The study took place at a not-for-profit health system offering a full continuum of care. The children's hospital is an urban, pediatric, academic hospital. This Magnet®-designated pediatric hospital has 200 pediatric physicians with 50 pediatric subspecialties and programs, a verified level 1 pediatric trauma center and designated level 4 neonatal intensive care unit. Our study did not include infants in the neonatal intensive care unit.
Retrospective Study Design & Method A retrospective chart review evaluated the association between PPUPET score and pressure ulcer development in pediatric patients admitted to our children's hospital during 2010–2011. This study was approved by our hospital's institutional review board (IRB) via expedited review, with waivers of consent, HIPAA authorization, parental permission, and assent. Sample and Sample Size Determination Our children's hospital averages 11,000 admissions per year. Between January 1, 2010 and December 31, 2011, a total of 77 pediatric patients developed a HAPU, according to quality improvement specialists' reports. A power analysis, performed to determine an adequate sample size for this study, determined that the required sample size to detect a medium effect for a chi square test was calculated as n = 220, while a large effect was n = 80, assuming a level of significance of p b 0.05 (alpha) and a power of β = 0.80. Thus, a randomized sample of 220 patients was planned from the population of admitted pediatric patients who did not develop a HAPU along with all (n = 77) pediatric patients who developed a HAPU. Of the 77 patients with a HAPU, 18 of these patients were excluded because they were admitted with a communityacquired pressure ulcer, admitted to neonatal intensive care, older than 18 years of age, or had no documented PPUPET scores. Of the 220 randomly selected patients (n = 110 from 2010 and n = 110 from 2011) drawn from the total number of admitted patients in the 2 year period who did not develop a HAPU, 6 were excluded for the same reasons. Thus, 59 HAPU patients' PPUPET assessments were compared to 214 non-HAPU patients' PPUPET assessments for a total of 273 patients, which fully powered the study.
D.J. Sterken et al. Data Collection PPUPET scores were collected from each subject's medical record. The admission PPUPET score was collected from every subject. The PPUPET score immediately prior to pressure ulcer development was collected on the HAPU patients. For non-HAPU patients, the last PPUPET score prior to discharge was collected. Data Analysis/Results The chi-square test was calculated to determine the predictive validity of the PPUPET and whether there was an association between PPUPET scores and patients' development of pressure ulcers. The chi-square is a two by two table consisting of four categories: at risk for pressure ulcers, not at risk for pressure ulcers, development of pressure ulcers, and no development of pressure ulcers. A chi-square analysis was performed for both the PPUPET score from admission and the PPUPET score upon discharge or prior to the development of a pressure ulcer. Of the 273 patients in the retrospective study, 59 developed a HAPU and 214 did not. The chi square test for the admission PPUPET score indicated that there was a significant association between the risk category indicated by the PPUPET score and the development of a pressure ulcer (χ2 (1) = 47.842, p b 0.001). In this sample, the sensitivity of the PPUPET was 74.58% (true positive result), and the specificity was 57.94% (true negative result). Likewise, the risk category indicated by the PPUPET score at discharge or prior to pressure ulcer development showed a significant association with the development of a pressure ulcer (X2 (1) = 90.708, p b 0.001). In this sample, the sensitivity was 76.27%, and the specificity was 75.70%. A logistic regression was also performed on the retrospective data, using the 9 components of the PPUPET as possible predictors for the development of a pressure ulcer. The logistic model correctly predicted whether or not a patient would develop a pressure ulcer 84% of the time. Considering only those patients with an “at risk” PPUPET score [total score ≥ 18; or nutrition score = 2; or any item score = 3], the percentage agreement with developing a pressure ulcer was 67.8%. Based on our study, there were four significant predictors for pressure ulcer development in children: mobility, skin conditions, friction/sheer, and external devices. Table 1 shows that higher scores in these 4subscales indicated greater odds of developing a pressure ulcer. Skin condition (OR = 2.269, 95% CI = [1.472, 3.495]) and external devices (OR = 1.587, 95% CI = [1.165, 2.161]) are items that are unique to the PPUPET tool. Mobility and friction/sheer are items included in the PPUPET, the Braden, and Braden Q tools.
Prospective Study The research questions for the prospective study were: 1. What is the predictive validity of the PPUPET as compared to the Braden Q (b 5 years of age Braden
Become the PPUPET Master
607
Table 1 Odds ratio (OR) estimates of developing a pressure ulcer by item. Item
Odds ratio
95% Confidence interval
Mobility Skin conditions Friction/Sheer External devices
2.986 2.269 2.154 1.587
2.067, 1.472, 1.262, 1.165,
4.132 3.495 3.678 2.161
Scale (≥ 5 years of age), or Glamorgan Scale in hospitalized pediatric patients? 2. Is the PPUPET more sensitive and specific than the Braden Q/Braden or Glamorgan Scale in predicting pressure ulcer risk in the hospitalized pediatric patient? Instruments For this study, the PPUPET was compared to three pressure ulcer risk assessment tools: the Braden, the Braden Q, and the Glamorgan Scale. The Braden, Braden Q, and PPUPET have been described previously in this article. The Glamorgan Scale was discovered by our research team as we updated our previous literature review. The Glamorgan Paediatric Pressure Ulcer Scale (Willock et al., 2008) was developed using expert opinion of factors considered to be relevant from a literature review and statistical analysis of patient data used to validate the scale (Anthony, Willock, & Baharestani, 2010). The scale identified 10 risk factors associated with pressure ulcers in children, and the risk score were adjusted so that the higher the total score, the greater the pressure ulcer risk (Willock et al., 2009). The scale was developed because current pediatric pressure ulcer risk assessment tools were merely modifications of adult scales for use with children. Willock et al. (2008) also identified medical equipment as a risk factor for the development of pressure ulcers. The Glamorgan scale has been compared to the Braden Q and found to have a higher level of predictability of risk of pressure ulcer (Anthony et al., 2010). Design and Method A prospective study comparing the PPUPET to the Braden/ Braden Q and Glamorgan was conducted on November 14, 2012 and February 13, 2013 during our hospital's usual quarterly HAPU audit. This study was approved by our hospital's IRB via expedited review, with a requirement for one parent's permission and assent from children ages 7–17 who have the maturity and cognitive ability to be reasonably consulted. Sample and Recruitment Study participants were recruited from a convenience sample of hospitalized patients meeting eligibility criteria during the quarterly HAPU prevalence audit. An informational letter was provided to the parent/guardian of each hospitalized patient the day before the audit. The letter described the purpose of the skin audit and the research study. The content of the letter provided an opportunity for
the parents to ask questions and identified that there was no risk to the child. This study required consent and HIPAA authorization from a parent or legal guardian for all participants and verbal assent from patients over the age of 7. There were 108 patients included as study participants. Data Collection The pediatric skin team consists of registered nurses who have received specialized training related to skin management and pressure ulcer prevention. Members of the skin team were supported by their nurse managers through dedicated paid time to participate on the team and in the study. Senior BSN students in a research practicum also assisted with data collection and analysis. Data collection was done by skin team members and nursing students during the quarterly HAPU audits. Before the audit began, all data collectors were oriented to the data collection instruments and completed a case study to assure inter-rater reliability for each tool. Data collectors introduced themselves to the child and parent and asked if they had read the parent letter delivered the previous day. If the parent was not aware, a letter was provided and explained. The patient and/or family members were informed that the HAPU audit was taking place and given a brief description of what would occur. Questions were answered about the audit and the study; then permission was requested to include the child in the study. If appropriate, the child was asked to assent to participation in the study. If the timing was inconvenient (e.g. the child was asleep), the nurses would come back at a later time that day to complete the audit and data collection. The patient was assigned a study ID number to protect privacy. This study ID number was written on all data collection forms. In consultation with the nurse and/or family members, direct observation, and review of the patient's medical record, the data collectors completed the appropriate instruments. One data collector completed a PPUPET risk assessment; one completed the age appropriate Braden or Braden Q; and a third completed the Glamorgan scale. An institutional HAPU audit tool was then used to collect actual pressure ulcer data and medical record information for each patient. The audit was not part of the research study. After consulting the assigned nurse of the patient, data collectors assessed the patient's skin, recording the findings on the audit tool and the other looking in the electronic medical record (EMR) for data. The assessment took 10–30 minutes depending on the patient. Any immediate care issues, such as a diaper change, were addressed as needed by the skin team member completing the audit. If a pressure ulcer was discovered, a wound care specialist was informed and assisted with documentation of ulcer staging, identified and assisted with the implementation of appropriate interventions, and notified the attending physician as per skin audit guidelines. The patient's medical record was reviewed for the current PPUPET score, a plan of care if indicated, and documentation of preventive interventions. Findings were
608
D.J. Sterken et al.
Hospital Acquired Pressure Ulcer Prevalence Percentage of Pressure Ulcers
12% 10% 8% 6% 4% 2%
Nov-13(2/62)
Aug-13(0/60)
Feb-13(0/69)
May-13(0/73)
Nov-12(2/58)
Aug-12(0/55)
May-12(1/55)
Feb-12(0/71)
Nov-11(0/51)
Aug-11(0/55)
Feb-11(0/53)
May-11(0/66)
Nov-10(0/54)
Aug-10(0/57)
May-10(1/50)
Feb-10 (0/84)
Nov-09 (2/67)
Nov-08 (2/48)
May-09 (1/52)
Nov-07 (2/60)
May-08 (2/57)
May-07 (0/54)
Jan-06 (6/74)
Nov-06 (2/55)
May-05 (7/66)
Nov-04 (1/40)
0%
Date of Audit Hospital Acquired Pressure Ulcer Prevalence
Figure 3
Hospital acquired pressure ulcer prevalence 2004–2013.
recorded on the audit tool. The only audit data recorded for this study was the presence or absence of pressure ulcers and the most recently documented PPUPET score.
Results The prospective study looked at 108 patients who had their HAPU risk assessed using the PPUPET, Glamorgan, and Braden or Braden Q. PPUPET scoring between the skin team and the bedside nurse indicated significant inter-rater reliability (κ = 0.718, p b 0.001). The level of agreement between the PPUPET and the Braden or Braden Q, when corrected for chance, was κ = 0.349, p b 0.001, which was statistically significant and a fair level of agreement. The level of agreement between the PPUPET and the Glamorgan was κ = 0.044, p = 0.281, which was not statistically significant and a slight level of agreement. None of the patients assessed in the two HAPU audits were found to have pressure ulcers; thus, the sensitivity (true positives) of the tools could not be calculated. The specificity (true negatives) of the PPUPET in the prospective study was 37.03%, which was less than the Braden or Braden Q specificity at 60.20% but greater than the Glamorgan specificity at 2.78%. The difference in specificity is most likely due to the way risk is defined in the PPUPET—either as total PPUPET score ≥ 18 or high risk in one of the subscales. The Braden, Braden Q, and Glamorgan instruments define risk based only on the total score.
Discussion/Conclusion/Clinical Implications Pressure ulcers are a problem in children. Staff nurses play a key role in identifying factors that place patients at risk
for the development of pressure ulcers. As a result of our efforts to identify risks for pressure ulcer development and implement interventions to prevent pressure ulcers, our institution has seen a decline in HAPU in our pediatric population (Figure 3). It is imperative that staff nurses have access to the following resources to prevent pressure ulcers in the pediatric population. First, nurses need education on the importance of accurate and thorough skin assessments to identify high risk patients on admission and throughout hospitalization. Second, creating a skin team with specialized education heightens awareness of issues directly related to the prevention of pressure ulcer development. Our skin team members serve as an ongoing resource for colleagues through consultation for skin care concerns (e.g., evaluating pressure ulcers, staging, and recommending product use) and product evaluation. Third, use of consistent language is important to describe injury to the skin secondary to pressure. Nursing staff are unlikely to see the difference between terms like “pressure ulcer” and “pressure injury.” Therefore, we consistently use the term “pressure ulcer” to describe any injury to the skin that is a result of pressure. Fourth, use of a consistent, reliable, and intuitive tool—like the PPUPET—provides a framework for clinical decision making related to pressure ulcer development. Screening tools should identify any risk factors that increase pressure to the skin; for example external medical devices. Scales which use low scores to demonstrate high risk and high scores to demonstrate low risk are confusing to bedside nurses and may
Become the PPUPET Master result in failure to initiate appropriate interventions. Lastly, providing bedside nurses with a plan of care (Figure 2) guides clinical decision-making and appropriate interventions to reduce the risk of pressure ulcer development, which promotes positive patient outcomes. The Braden and Braden Q Scales are the most widely recognized and used pressure ulcer risk assessment scales. The Braden scale is not intended for use in children, but the Braden Q was modified for use in children from 3 weeks to 8 years of age. Some children's hospitals are using the Braden Q for children of all ages. In our studies, the PPUPET was used in children of all ages. Our study indicates that the level of agreement between the PPUPET and the Braden or Braden Q was statistically significant. In addition, the PPUPET identifies two unique items (skin condition and external medical devices) that are significant predictors of risk for pressure ulcer development in the pediatric patient. A limitation of our study was that it was conducted with patients from one institution. Over the past 10 years, our hospital has increased its emphasis on identifying children at risk for pressure ulcers and implementing appropriate interventions. As a result, our experience with the PPUPET may be positively biased due to the dedicated resources of an active skin care team along with ongoing quarterly education. The PPUPET scores on admission and discharge significantly predicted (p b 0.001) the risk of children who developed pressure ulcers in the retrospective study. The sensitivity (true positive result) at admission was 74.58% (n = 44) and at discharge was 76.27% (n = 45). Since no children in the prospective study had HAPU's, the sensitivity could not be calculated. The PPUPET is highly sensitive even at admission for predicting patients at high risk. The specificity (true negative result) at admission was 57.94% (n = 124) and at discharge was 75.70% (n = 162) in the retrospective study. These results indicate that the PPUPET was moderately specific at admission and highly specific at discharge at predicting children who are not likely to develop pressure ulcers. In the prospective study, however, the specificity was 37.03%, with only 40 out of 108 children scoring not at risk for pressure ulcer development. This can be explained by the way risk is defined in the PPUPET as either a total score ≥ 18 or high risk (2 or 3) in any subscale score. If only a total PPUPET score had been used to classify risk (similar to the Braden, Braden Q, and Glamorgan tools), 99 out of 108 children (91.67%) would have been identified as not at risk. Instead, the PPUPET classified 68 children as being at risk, yet none of them were found to have a HAPU in the audit. This may be attributed to the nurses' interventions to prevent pressure ulcer development. We believe the PPUPET can be used for evaluating pressure ulcer risk in children from birth to age 18.
609 Continued research is needed to evaluate the reliability, sensitivity, specificity, and predictive validity of the PPUPET. We recommend a multi-site study for further validation of the PPUPET in geographically diverse populations.
Acknowledgments We would like to express our gratitude to those who helped in the development and validation of the PPUPET: Marcia MacGeorge, Michelle Oleniczak, Kristi Jackson, Jessica McClusky, Nichole Fieldhouse, Jennifer Fox, Angela Snellen, Rebecca DeVries, Sarah Butterfield, Amy Overway, Karen Sweet, Todd Nickoles, Tammy Raterink, Nancy Doyle, Denise Katerberg, Amanda Vandenberg, Cathie Vandersteen, Alesia Derks, Catherine Keezel, Emily Carlson, Kathleen Puff, Andrea Lynn Rosendahl, Kaylee Marcinkus, Barbara Vincensi, Donna Garrett, Jenna Dietsch, and Jody Kinney. Authors have no conflict of interest.
References Anthony, D., Willock, J., & Baharestani, M. (2010). A comparison of Braden Q, Garvin, and Glamorgan risk assessment scales in paediatrics. Journal of Tissue Viability, 19, 98–105. Bergstrom, N., Braden, B. J., Laguzza, A., & Holman, V. (1987). The Braden Scale for predicting pressure sore risk. Nursing Research, 36, 206–210. Bergstrom, N., Demuth, P. J., & Braden, B. J. (1987). A clinical trial of the Braden Scale for predicting pressure sore risk. Nursing Clinics of North America, 22, 417–428. Bernabe, K. Q. (2012). Pressure ulcers in the pediatric patient. Current Opinion in Pediatrics, 24, 352–356. Curley, M., Razmus, I., Roberts, K., & Wypij, D. (2003). Predicting pressure ulcer risk inpediatric patients: The Braden Q Scale. Journal of Eastern Nursing Research Society, 22–33. European Pressure Ulcer Advisory Panel, & National Pressure Ulcer Advisory Panel (2009). Prevention and treatment of pressure ulcers: Quick reference guide. Washington DC: National Pressure Ulcer Advisory Panel. Kottner, J., Wilborn, D., & Dassen, T. (2010). Frequency of pressure ulcers in the paediatric population: A literature review and new empirical data. International Journal of Nursing Studies, 47, 1330–1340. MedCalc Software (2014). MedCalc Easy-to-Use Statistical Software: Diagnostic test calculator (Last modified 05-22-14) [Software]. Available from. http://www.medcalc.org/calc/diagnostic_test.php Noonan, C., Quigley, S., & Curley, M. (2006). Skin integrity in hospitalized infants and children. Journal of Pediatric Nursing, 21, 445–453. Parnham, A. (2012). Pressure ulcer risk assessment and prevention in children. Nursing Children and Young People, 24, 24–29. Pediatric Affinity Group (2014). How-to-guide pediatric supplement: Preventing pressure ulcers (Last updated 02–01–14). Available from. http://www.nichq.org/pdf/FINALPressureUlcers. pdf Polit, D. F., & Beck, C. T. (2013). Essentials of nursing research: Appraising evidence for nursing practice (8th ed.). Philadelphia, PA: Lippincott Williams & Wilkins. Quigley, S., & Curley, M. (1996). Skin integrity in the pediatric population: Preventing and managing pressure ulcers. Journal for Specialists in Pediatric Nursing, 1, 7–18.
610 Sim, J., & Wright, C. C. (2005). The Kappa statistic in reliability studies: Use, interpretation, and sample size requirements. Physical Therapy, 85, 257–268. VanderKooi, M., Blackport, C., & Vander Laan, K. J. (2011). Nurses’ information flow and communication: A comparison of paper to electronic Kardex – Part 3 survey. Poster presented at Spectrum Health Nursing Research Traveling Poster Exhibit, Grand Rapids, Michigan. Waterlow, J. (1997). Pressure sore risk assessment in children. Paediatric Nursing, 9, 21–24.
D.J. Sterken et al. Willock, J., Anthony, D., & Richardson, J. (2008). Inter-rater reliability of the Glamorgan Pediatric pressure ulcer risk assessment scale. Pediatric Nursing, 20, 14–19. Willock, J., Baharestani, M. M., & Anthony, D. (2009). The development of the Glamorgan paediatric pressure ulcer risk assessment scale. Journal of Wound Care, 18, 17–21. Willock, J., Harris, C., Harrison, J., & Poole, C. (2005). Identifying the characteristics of children with pressure ulcers. Nursing Times, 101, 40–43.
Become the PPUPET Master: Mastering Pressure Ulcer Risk Assessment With the Pediatric Pressure Ulcer Prediction and Evaluation Tool (PPUPET)
David J. Sterken MN, CPNP a,⁎, JoAnn Mooney BSN, RN CPN b , Diana Ropele MSN, RN, CCRN c , Alysha Kett BS, MS d , Karen J. Vander Laan PhD, MSN, RN e a
Pediatric Services, Helen DeVos Children's Hospital, Grand Rapids, MI Quality Improvement, Helen DeVos Children's Hospital, Grand Rapids, MI c Pediatric Trauma Services, Helen DeVos Children's Hospital, Grand Rapids, MI d Spectrum Health, Grand Rapids, MI e Spectrum Health Hospitals, Center for Nursing Practice & Development, Grand Rapids, MI b
Received 24 February 2014; revised 2 October 2014; accepted 7 October 2014
Key words: Hospital acquired pressure ulcer (HAPU); Pressure ulcer; Pediatric pressure ulcers; PPUPET
Hospital acquired pressure ulcers (HAPU) are serious, debilitating, and preventable complications in all inpatient populations. Despite evidence of the development of pressure ulcers in the pediatric population, minimal research has been done. Based on observations gathered during quarterly HAPU audits, bedside nursing staff recognized trends in pressure ulcer locations that were not captured using current pressure ulcer risk assessment tools. Together, bedside nurses and nursing leadership created and conducted multiple research studies to investigate the validity and reliability of the Pediatric Pressure Ulcer Prediction and Evaluation Tool (PPUPET). © 2015 Elsevier Inc. All rights reserved.
NOSOCOMIAL PRESSURE ULCERS are a serious, debilitating, and preventable complication in all inpatient populations. Similar to adult patients, acutely ill infants and children are at risk for pressure ulcers (Bernabe, 2012; Kottner, Wilborn, & Dassen, 2010; Parnham, 2012; Quigley & Curley, 1996). However, because of the anatomical and physiological differences in infants and children, the most common sites for pressure ulcer development are different than the adult population. For example, the head makes up a greater proportion of the total body weight and surface; thus, the occipital region of the scalp is the most common site of ulceration for infants and children (Willock, Harris, Harrison, & Poole, 2005). Infants and children are also prone to develop pressure ulcers on other bony prominences as well as any area where external medical devices (such as ⁎ Corresponding author: David J. Sterken, MN, CPNP. E-mail address: [email protected]. http://dx.doi.org/10.1016/j.pedn.2014.10.004 0882-5963/© 2015 Elsevier Inc. All rights reserved.
pulse oximetry probes, oxygen tubing, intravenous hubs, braces or other tubes) are placed (Kottner et al., 2010; Waterlow, 1997; Willock, Baharestani, & Anthony, 2009). Some conditions that place children at greater risk for pressure ulcer development include extracorporeal membrane oxygenation (ECMO), paraplegia, myelomeningocele, large head size, kyphosis, developmental delay, and chronic fecal and urinary soiling (Noonan, Quigley, & Curley, 2006). Despite evidence of the development of pressure ulcers in the pediatric population, relatively minimal research has been done to develop pediatric pressure ulcer risk assessment tools. The purpose of this article is to introduce and report testing of the Pediatric Pressure Ulcer Prediction and Evaluation Tool (PPUPET), a pediatric-focused skin risk assessment instrument developed by bedside nurse leaders in response to finding pressure ulcers in our patient population. The use of standardized tools in nursing promotes consistency among caregivers and facilitates communication
Become the PPUPET Master regarding evaluation and management of clinical decisions and treatments for various clinical conditions. In pediatrics, however, standardized tools are often modifications of adult tools and do not always address the unique considerations of the pediatric patient. Standardized tools must also be re-evaluated and revised periodically as healthcare changes and the language used to describe clinical conditions becomes more uniform. Standardized tools must be easy to use and intuitive for the bedside nurse; otherwise, the tool can add burden which takes away from patient care. Evaluation testing of standardized tools is also important to ensure that the tool consistently measures the clinical condition and that appropriate assessment criteria are being considered. Inter-rater testing, which assesses the degree to which different raters independently concur in their observations of what is being measured, is used to report a tool's reliability (Polit & Beck, 2013). Validity—the degree to which an instrument measures what it is intended to measure—can be established by comparing the items in a tool to another reliable, valid instrument measuring similar concepts (Polit & Beck, 2013). When a tool is used in clinical practice at a time when outcome conditions can also be assessed, calculations of sensitivity—the tool's ability to identify patients who have a true positive condition—and specificity—the tool's ability to identify patients who have a true negative condition—help clinicians appraise the usefulness of the standardized tool (MedCalc Software, 2014). For the PPUPET, sensitivity measures the proportion of children rated at risk who actually have a pressure ulcer, while specificity measures the proportion of children rated not at risk who actually have no pressure ulcers. Lastly, positive and negative predictive value can also be calculated for an instrument (MedCalc Software, 2014). For the PPUPET, the positive predictive value measures the probability that the risk for pressure ulcer development is present when the child is assessed as being at risk for pressure ulcer development. Negative predictive value measures the probability that the risk for pressure ulcer development is not present when the child is assessed as not being at risk.
Purpose and Background The purpose of this article is to report the results of recent retrospective and prospective studies of the PPUPET which have evaluated its usefulness as a risk assessment tool for the development of pressure ulcers in the pediatric population. As background, we will describe the development and initial reliability/validity testing of the PPUPET and then discuss the more recent studies.
Bringing the PPUPET to Life In the fall of 2004, members of the children's hospital nursing leadership team began a conversation regarding the prevalence of pressure ulcers in our pediatric patient population. Our adult hospital colleagues already had been performing hospital acquired pressure ulcer (HAPU) audits to develop strategies for the prevention and treatment of
599 pressure ulcers. In order to establish a baseline for prevalence of pressure ulcers, our first skin audit was completed in November 2004. A complete head to toe skin assessment was conducted on 49 hospitalized pediatric patients on a single day. Results of our audit demonstrated a prevalence of pressure ulcers in 9% of these patients. In May 2005, the next skin audit was performed on 59 patients. Results of the audit demonstrated that 15% of the patients had a pressure ulcer. At this time the nursing leadership team realized that pressure ulcers were indeed a problem at the children's hospital and began to formulate a strategy to address the issue. Nursing leadership recruited bedside nurses from each clinical unit to form a team to address the prevalence of pressure ulcers in our pediatric patient population. Members of the “Skin Team” were assigned to conduct biannual skin audits. Following the January 2006 skin audit of 74 pediatric patients—which yielded an alarming 11% prevalence of stage 1 (n = 8) and 1% stage 2 (n = 1) pressure ulcers1 and several patients having N 1 pressure ulcer—the skin team identified the following concerns: • No standardized tool was being used to identify patients at risk for pressure ulcer development. • No pediatric plan of care had been developed for the prevention and management of pressure ulcers in children. • The nursing Kardex2 did not reflect nursing orders for turning and basic skin care. • Nurses were unaware that they could initiate nursing orders/interventions to prevent the development of pressure ulcers. • Documentation regarding nursing interventions (e.g., turning and positioning) for the prevention of pressure ulcers was lacking or inconsistent. • Nurses were unaware of where or how to document positioning in new electronic nursing documentation. • Care related to devices (e.g., repositioning pulse oximetry probes) was inconsistently documented.
Based on the identified concerns, the team decided to review standardized pressure ulcer risk assessment tools and develop a standard of practice for assessment and intervention to prevent pediatric pressure ulcers. Starting in March 2006, skin team members began reviewing the existing evidence. We first reviewed the Braden Scale for Predicting Pressure Sore Risk (Bergstrom, Braden, Laguzza, & Holman, 1987) [Braden], which was the tool used by our 1
Pressure ulcer staging is based on the National Pressure Ulcer Advisory Panel's (NPUAP) definitions of stages 1 through 4. Stage 1 is the least severe (non-blanchable erythema), and stage 2 is a partial-thickness ulcer. The current version of the Quick Reference Guide (European Pressure Ulcer Advisory Panel and National Pressure Ulcer Advisory Panel, 2009), now includes stages 1 to 4 and deep tissue injury and unstageable categories. 2 The paper Kardex was a primary tool utilized by bedside nursing staff to communicate pertinent patient-specific information, direct nursing care based on physician orders, facilitate nurse-to-nurse communication, and track patient significant events (VanderKooi, Blackport, & Vander Laan, 2011).
600 colleagues caring for adult patients. The scale has 6 subscales: sensory perception, moisture, activity, mobility, nutrition, and friction & shear. Each subscale is mutually exclusive. Each item of the subscale has a minimum score of 1 (more risk) and a maximum score of 3 or 4 (less risk). The range of scores is 6–23. In the acutely ill adult population cut-off points of 16 to 18 are used to identify patients at risk for developing pressure ulcers. At a score of 16 the Braden Scale has a sensitivity of 0.83 and specificity of 0.64 in predicting pressure ulcer formation (Bergstrom, Demuth, & Braden, 1987). Next we looked at the Braden Q Scale for Predicting Pressure Ulcer Risk (Quigley & Curley, 1996) [Braden Q], which was validated for use in pediatric patients from 3 weeks to 8 years of age (Curley, Razmus, Roberts, & Wypij, 2003). They made changes to the original scale related to developmental needs and treatment interventions common to the pediatric patient and added a seventh subscale for tissue perfusion and oxygenation. The minimal score for each subscale is 1 (more risk), and the maximum score is 4 (less risk). Potential scores range from 7 to 28 points, the lower the score the higher the risk for ulcer development. Content validity was established by a group of pediatric nurses with special interest in skin issues. Confidence intervals indicated that children scoring b 23 were at moderate or high risk for pressure ulcers (Quigley & Curley, 1996). An initial study to establish the predictive validity of the tool revealed that at a score of 16 the sensitivity was 0.88 and the specificity was 0.58 (Curley et al., 2003). Concepts from these validated tools were recognized as important to pediatric pressure ulcer risk assessment. Members of the committee felt that the Braden and Braden Q did not identify some key risk factors unique to our population and confirmed by our skin audits. First, a compromised skin condition was not identified as a risk factor for pressure ulcer development. Second, external medical devices—such as pulse oximetry probes, intravenous hubs/protective devices, and orthopedic equipment— were not recognized as significant causes of pressure ulcers. In addition, use of the Braden and Braden Q required two different tools based on the age of the patient. Lastly, nursing staff indicated that the Braden and Braden Q tools were not intuitive in scoring, as the higher the score the lower the risk of developing a pressure ulcer. We then explored the possibility of adding additional subscales to the Braden Q in a conversation with Dr. Curley (personal communication, October 2006). However, at that time, modification of the Braden Q was not an option due to copyright restrictions. Subsequently, a decision was made to develop a tool specific for our population, using valid and reliable concepts related to pressure ulcer risk from the literature, our institution's sources of pressure ulcers, and recommendations from clinical nursing staff. The Pediatric Pressure Ulcer Prediction and Evaluation Tool (PPUPET) was created in 2007 (Figure 1) for use with children from birth to 18 years, with the exception of premature infants cared for in the neonatal intensive care unit. The PPUPET
D.J. Sterken et al. covers 9 subscales that are mutually exclusive. The subscales include mobility, activity, sensory perception, moisture, external medical devices, friction/shear, tissue perfusion & oxygenation, skin condition and nutrition. Each item of the subscale has a minimum score of 1 (less risk) and a maximum score of 2 or 3 (more risk). The range of scores is 9–26. A score of 18–26 indicates that the patient is at risk for developing a pressure ulcer. In addition to the total score, a score of 3 in any subscale or a 2 in the nutrition subscale also puts a patient at risk. All children at risk of pressure ulcer development require implementation of a plan of care for the prevention of skin breakdown. The PPUPET included external medical devices and skin condition as two subscales not previously identified in the literature. To make the tool intuitive, higher scores reflected greater risk of pressure ulcer development. The tool was piloted and adjusted to incorporate the clinical nursing staff's suggestions regarding reference text for each subscale. In conjunction with the PPUPET, a plan of care (Figure 2) was created to identify specific key interventions to be implemented for risks identified by high subscale scores. Education was then provided to all staff regarding the risks of development of pressure ulcers, especially those related to medical devices; interventions for prevention of pressure ulcers, including repositioning of medical devices if possible; and instructions for documentation of interventions,
Testing the PPUPET's Performance Next the skin team, led by a clinical nurse specialist, designed a validation study to compare the PPUPET to current validated scales identified in the literature and to document inter-rater reliability. For our study we used the Braden for patients age 5 years and older and the Braden Q for patients under 5 years of age, as provided in the Institute for Healthcare Improvement's How-to-guide Pediatric Supplement Preventing Pressure Ulcers (Pediatric Affinity Group, last updated 02-01-2014). In November 2008, the pressure ulcer risk prediction from the items assessed in the PPUPET was compared to the Braden or Braden Q scales during that quarter's HAPU prevalence audit of 48 pediatric patients. The PPUPET identified 7 patients (14.7%) as high risk for pressure ulcer development; 2 of these patients (4.2%) were found to actually have pressure ulcers. This computes to a sensitivity of 100.00% (95% CI 19.29, 100.00), specificity of 89.13% (95% CI 76.48, 96.33), a positive predictive value of 28.51% (95% CI 4.52, 70.73), and a negative predictive value of 100.00% (95% CI 91.31, 100.00). The percentage of agreement between the PPUPET and the Braden Q (for patients b 5 years of age) was 97% (n = 30) with a Kappa3 of 3 The Kappa statistic corrects for the number of times agreement might be expected to happen by chance, so it is a more accurate reflection of the actual (true) agreement (Sim & Wright, 2005).
Become the PPUPET Master
Figure 1
601
Pediatric pressure ulcer prediction and evaluation tool (PPUPET).
0.650. The percentage of agreement between the PPUPET and the Braden (for patients ≥ 5 years of age) was 83% (n = 18) with a Kappa of 0.557. Scoring of patients as high risk on the PPUPET was significantly related to scoring as high risk on the Braden Q (χ2 (1) = 14.483, p b 0.001) and the Braden (χ2 (1) = 5.716, p = 0.017). Inter-rater reliability was tested by comparing staff nurses' most recent PPUPET scores to Skin Champion's PPUPET scores on the day of the hospital skin audit. There was a significant difference (t (43) = − 2.922, p = 0.005) between the staff nurses' PPUPET score (mean = 12.78,
std dev = 3.649) and the Skin Champions' PPUPET score (mean = 13.49, std dev = 3.764), which led to further staff education and revisions of the reference text. The current version of the PPUPET was adopted as a standard nursing assessment in 2009. Studies to Validate the PPUPET In 2012, two studies—one retrospective and one prospective—were designed to validate the PPUPET. The purpose of the retrospective study was to review the medical
602
Prevention of skin breakdown interdisciplinary plan of care (Page 1 of 4).
D.J. Sterken et al.
Figure 2
603
Figure 2 (continued).
Become the PPUPET Master
D.J. Sterken et al.
Figure 2 (continued).
604
605
Figure 2 (continued).
Become the PPUPET Master
606 records of pediatric patients hospitalized between January 1, 2010 to December 31, 2012 who had been assessed using the PPUPET to evaluate the association between PPUPET score and pressure ulcer development. The purpose of the prospective study was to compare the predictive validity of the PPUPET to the predictive validity of the Braden or Braden Q and to a new tool, the Glamorgan (Willock, Anthony, & Richardson, 2008).
Methodology Setting The study took place at a not-for-profit health system offering a full continuum of care. The children's hospital is an urban, pediatric, academic hospital. This Magnet®-designated pediatric hospital has 200 pediatric physicians with 50 pediatric subspecialties and programs, a verified level 1 pediatric trauma center and designated level 4 neonatal intensive care unit. Our study did not include infants in the neonatal intensive care unit.
Retrospective Study Design & Method A retrospective chart review evaluated the association between PPUPET score and pressure ulcer development in pediatric patients admitted to our children's hospital during 2010–2011. This study was approved by our hospital's institutional review board (IRB) via expedited review, with waivers of consent, HIPAA authorization, parental permission, and assent. Sample and Sample Size Determination Our children's hospital averages 11,000 admissions per year. Between January 1, 2010 and December 31, 2011, a total of 77 pediatric patients developed a HAPU, according to quality improvement specialists' reports. A power analysis, performed to determine an adequate sample size for this study, determined that the required sample size to detect a medium effect for a chi square test was calculated as n = 220, while a large effect was n = 80, assuming a level of significance of p b 0.05 (alpha) and a power of β = 0.80. Thus, a randomized sample of 220 patients was planned from the population of admitted pediatric patients who did not develop a HAPU along with all (n = 77) pediatric patients who developed a HAPU. Of the 77 patients with a HAPU, 18 of these patients were excluded because they were admitted with a communityacquired pressure ulcer, admitted to neonatal intensive care, older than 18 years of age, or had no documented PPUPET scores. Of the 220 randomly selected patients (n = 110 from 2010 and n = 110 from 2011) drawn from the total number of admitted patients in the 2 year period who did not develop a HAPU, 6 were excluded for the same reasons. Thus, 59 HAPU patients' PPUPET assessments were compared to 214 non-HAPU patients' PPUPET assessments for a total of 273 patients, which fully powered the study.
D.J. Sterken et al. Data Collection PPUPET scores were collected from each subject's medical record. The admission PPUPET score was collected from every subject. The PPUPET score immediately prior to pressure ulcer development was collected on the HAPU patients. For non-HAPU patients, the last PPUPET score prior to discharge was collected. Data Analysis/Results The chi-square test was calculated to determine the predictive validity of the PPUPET and whether there was an association between PPUPET scores and patients' development of pressure ulcers. The chi-square is a two by two table consisting of four categories: at risk for pressure ulcers, not at risk for pressure ulcers, development of pressure ulcers, and no development of pressure ulcers. A chi-square analysis was performed for both the PPUPET score from admission and the PPUPET score upon discharge or prior to the development of a pressure ulcer. Of the 273 patients in the retrospective study, 59 developed a HAPU and 214 did not. The chi square test for the admission PPUPET score indicated that there was a significant association between the risk category indicated by the PPUPET score and the development of a pressure ulcer (χ2 (1) = 47.842, p b 0.001). In this sample, the sensitivity of the PPUPET was 74.58% (true positive result), and the specificity was 57.94% (true negative result). Likewise, the risk category indicated by the PPUPET score at discharge or prior to pressure ulcer development showed a significant association with the development of a pressure ulcer (X2 (1) = 90.708, p b 0.001). In this sample, the sensitivity was 76.27%, and the specificity was 75.70%. A logistic regression was also performed on the retrospective data, using the 9 components of the PPUPET as possible predictors for the development of a pressure ulcer. The logistic model correctly predicted whether or not a patient would develop a pressure ulcer 84% of the time. Considering only those patients with an “at risk” PPUPET score [total score ≥ 18; or nutrition score = 2; or any item score = 3], the percentage agreement with developing a pressure ulcer was 67.8%. Based on our study, there were four significant predictors for pressure ulcer development in children: mobility, skin conditions, friction/sheer, and external devices. Table 1 shows that higher scores in these 4subscales indicated greater odds of developing a pressure ulcer. Skin condition (OR = 2.269, 95% CI = [1.472, 3.495]) and external devices (OR = 1.587, 95% CI = [1.165, 2.161]) are items that are unique to the PPUPET tool. Mobility and friction/sheer are items included in the PPUPET, the Braden, and Braden Q tools.
Prospective Study The research questions for the prospective study were: 1. What is the predictive validity of the PPUPET as compared to the Braden Q (b 5 years of age Braden
Become the PPUPET Master
607
Table 1 Odds ratio (OR) estimates of developing a pressure ulcer by item. Item
Odds ratio
95% Confidence interval
Mobility Skin conditions Friction/Sheer External devices
2.986 2.269 2.154 1.587
2.067, 1.472, 1.262, 1.165,
4.132 3.495 3.678 2.161
Scale (≥ 5 years of age), or Glamorgan Scale in hospitalized pediatric patients? 2. Is the PPUPET more sensitive and specific than the Braden Q/Braden or Glamorgan Scale in predicting pressure ulcer risk in the hospitalized pediatric patient? Instruments For this study, the PPUPET was compared to three pressure ulcer risk assessment tools: the Braden, the Braden Q, and the Glamorgan Scale. The Braden, Braden Q, and PPUPET have been described previously in this article. The Glamorgan Scale was discovered by our research team as we updated our previous literature review. The Glamorgan Paediatric Pressure Ulcer Scale (Willock et al., 2008) was developed using expert opinion of factors considered to be relevant from a literature review and statistical analysis of patient data used to validate the scale (Anthony, Willock, & Baharestani, 2010). The scale identified 10 risk factors associated with pressure ulcers in children, and the risk score were adjusted so that the higher the total score, the greater the pressure ulcer risk (Willock et al., 2009). The scale was developed because current pediatric pressure ulcer risk assessment tools were merely modifications of adult scales for use with children. Willock et al. (2008) also identified medical equipment as a risk factor for the development of pressure ulcers. The Glamorgan scale has been compared to the Braden Q and found to have a higher level of predictability of risk of pressure ulcer (Anthony et al., 2010). Design and Method A prospective study comparing the PPUPET to the Braden/ Braden Q and Glamorgan was conducted on November 14, 2012 and February 13, 2013 during our hospital's usual quarterly HAPU audit. This study was approved by our hospital's IRB via expedited review, with a requirement for one parent's permission and assent from children ages 7–17 who have the maturity and cognitive ability to be reasonably consulted. Sample and Recruitment Study participants were recruited from a convenience sample of hospitalized patients meeting eligibility criteria during the quarterly HAPU prevalence audit. An informational letter was provided to the parent/guardian of each hospitalized patient the day before the audit. The letter described the purpose of the skin audit and the research study. The content of the letter provided an opportunity for
the parents to ask questions and identified that there was no risk to the child. This study required consent and HIPAA authorization from a parent or legal guardian for all participants and verbal assent from patients over the age of 7. There were 108 patients included as study participants. Data Collection The pediatric skin team consists of registered nurses who have received specialized training related to skin management and pressure ulcer prevention. Members of the skin team were supported by their nurse managers through dedicated paid time to participate on the team and in the study. Senior BSN students in a research practicum also assisted with data collection and analysis. Data collection was done by skin team members and nursing students during the quarterly HAPU audits. Before the audit began, all data collectors were oriented to the data collection instruments and completed a case study to assure inter-rater reliability for each tool. Data collectors introduced themselves to the child and parent and asked if they had read the parent letter delivered the previous day. If the parent was not aware, a letter was provided and explained. The patient and/or family members were informed that the HAPU audit was taking place and given a brief description of what would occur. Questions were answered about the audit and the study; then permission was requested to include the child in the study. If appropriate, the child was asked to assent to participation in the study. If the timing was inconvenient (e.g. the child was asleep), the nurses would come back at a later time that day to complete the audit and data collection. The patient was assigned a study ID number to protect privacy. This study ID number was written on all data collection forms. In consultation with the nurse and/or family members, direct observation, and review of the patient's medical record, the data collectors completed the appropriate instruments. One data collector completed a PPUPET risk assessment; one completed the age appropriate Braden or Braden Q; and a third completed the Glamorgan scale. An institutional HAPU audit tool was then used to collect actual pressure ulcer data and medical record information for each patient. The audit was not part of the research study. After consulting the assigned nurse of the patient, data collectors assessed the patient's skin, recording the findings on the audit tool and the other looking in the electronic medical record (EMR) for data. The assessment took 10–30 minutes depending on the patient. Any immediate care issues, such as a diaper change, were addressed as needed by the skin team member completing the audit. If a pressure ulcer was discovered, a wound care specialist was informed and assisted with documentation of ulcer staging, identified and assisted with the implementation of appropriate interventions, and notified the attending physician as per skin audit guidelines. The patient's medical record was reviewed for the current PPUPET score, a plan of care if indicated, and documentation of preventive interventions. Findings were
608
D.J. Sterken et al.
Hospital Acquired Pressure Ulcer Prevalence Percentage of Pressure Ulcers
12% 10% 8% 6% 4% 2%
Nov-13(2/62)
Aug-13(0/60)
Feb-13(0/69)
May-13(0/73)
Nov-12(2/58)
Aug-12(0/55)
May-12(1/55)
Feb-12(0/71)
Nov-11(0/51)
Aug-11(0/55)
Feb-11(0/53)
May-11(0/66)
Nov-10(0/54)
Aug-10(0/57)
May-10(1/50)
Feb-10 (0/84)
Nov-09 (2/67)
Nov-08 (2/48)
May-09 (1/52)
Nov-07 (2/60)
May-08 (2/57)
May-07 (0/54)
Jan-06 (6/74)
Nov-06 (2/55)
May-05 (7/66)
Nov-04 (1/40)
0%
Date of Audit Hospital Acquired Pressure Ulcer Prevalence
Figure 3
Hospital acquired pressure ulcer prevalence 2004–2013.
recorded on the audit tool. The only audit data recorded for this study was the presence or absence of pressure ulcers and the most recently documented PPUPET score.
Results The prospective study looked at 108 patients who had their HAPU risk assessed using the PPUPET, Glamorgan, and Braden or Braden Q. PPUPET scoring between the skin team and the bedside nurse indicated significant inter-rater reliability (κ = 0.718, p b 0.001). The level of agreement between the PPUPET and the Braden or Braden Q, when corrected for chance, was κ = 0.349, p b 0.001, which was statistically significant and a fair level of agreement. The level of agreement between the PPUPET and the Glamorgan was κ = 0.044, p = 0.281, which was not statistically significant and a slight level of agreement. None of the patients assessed in the two HAPU audits were found to have pressure ulcers; thus, the sensitivity (true positives) of the tools could not be calculated. The specificity (true negatives) of the PPUPET in the prospective study was 37.03%, which was less than the Braden or Braden Q specificity at 60.20% but greater than the Glamorgan specificity at 2.78%. The difference in specificity is most likely due to the way risk is defined in the PPUPET—either as total PPUPET score ≥ 18 or high risk in one of the subscales. The Braden, Braden Q, and Glamorgan instruments define risk based only on the total score.
Discussion/Conclusion/Clinical Implications Pressure ulcers are a problem in children. Staff nurses play a key role in identifying factors that place patients at risk
for the development of pressure ulcers. As a result of our efforts to identify risks for pressure ulcer development and implement interventions to prevent pressure ulcers, our institution has seen a decline in HAPU in our pediatric population (Figure 3). It is imperative that staff nurses have access to the following resources to prevent pressure ulcers in the pediatric population. First, nurses need education on the importance of accurate and thorough skin assessments to identify high risk patients on admission and throughout hospitalization. Second, creating a skin team with specialized education heightens awareness of issues directly related to the prevention of pressure ulcer development. Our skin team members serve as an ongoing resource for colleagues through consultation for skin care concerns (e.g., evaluating pressure ulcers, staging, and recommending product use) and product evaluation. Third, use of consistent language is important to describe injury to the skin secondary to pressure. Nursing staff are unlikely to see the difference between terms like “pressure ulcer” and “pressure injury.” Therefore, we consistently use the term “pressure ulcer” to describe any injury to the skin that is a result of pressure. Fourth, use of a consistent, reliable, and intuitive tool—like the PPUPET—provides a framework for clinical decision making related to pressure ulcer development. Screening tools should identify any risk factors that increase pressure to the skin; for example external medical devices. Scales which use low scores to demonstrate high risk and high scores to demonstrate low risk are confusing to bedside nurses and may
Become the PPUPET Master result in failure to initiate appropriate interventions. Lastly, providing bedside nurses with a plan of care (Figure 2) guides clinical decision-making and appropriate interventions to reduce the risk of pressure ulcer development, which promotes positive patient outcomes. The Braden and Braden Q Scales are the most widely recognized and used pressure ulcer risk assessment scales. The Braden scale is not intended for use in children, but the Braden Q was modified for use in children from 3 weeks to 8 years of age. Some children's hospitals are using the Braden Q for children of all ages. In our studies, the PPUPET was used in children of all ages. Our study indicates that the level of agreement between the PPUPET and the Braden or Braden Q was statistically significant. In addition, the PPUPET identifies two unique items (skin condition and external medical devices) that are significant predictors of risk for pressure ulcer development in the pediatric patient. A limitation of our study was that it was conducted with patients from one institution. Over the past 10 years, our hospital has increased its emphasis on identifying children at risk for pressure ulcers and implementing appropriate interventions. As a result, our experience with the PPUPET may be positively biased due to the dedicated resources of an active skin care team along with ongoing quarterly education. The PPUPET scores on admission and discharge significantly predicted (p b 0.001) the risk of children who developed pressure ulcers in the retrospective study. The sensitivity (true positive result) at admission was 74.58% (n = 44) and at discharge was 76.27% (n = 45). Since no children in the prospective study had HAPU's, the sensitivity could not be calculated. The PPUPET is highly sensitive even at admission for predicting patients at high risk. The specificity (true negative result) at admission was 57.94% (n = 124) and at discharge was 75.70% (n = 162) in the retrospective study. These results indicate that the PPUPET was moderately specific at admission and highly specific at discharge at predicting children who are not likely to develop pressure ulcers. In the prospective study, however, the specificity was 37.03%, with only 40 out of 108 children scoring not at risk for pressure ulcer development. This can be explained by the way risk is defined in the PPUPET as either a total score ≥ 18 or high risk (2 or 3) in any subscale score. If only a total PPUPET score had been used to classify risk (similar to the Braden, Braden Q, and Glamorgan tools), 99 out of 108 children (91.67%) would have been identified as not at risk. Instead, the PPUPET classified 68 children as being at risk, yet none of them were found to have a HAPU in the audit. This may be attributed to the nurses' interventions to prevent pressure ulcer development. We believe the PPUPET can be used for evaluating pressure ulcer risk in children from birth to age 18.
609 Continued research is needed to evaluate the reliability, sensitivity, specificity, and predictive validity of the PPUPET. We recommend a multi-site study for further validation of the PPUPET in geographically diverse populations.
Acknowledgments We would like to express our gratitude to those who helped in the development and validation of the PPUPET: Marcia MacGeorge, Michelle Oleniczak, Kristi Jackson, Jessica McClusky, Nichole Fieldhouse, Jennifer Fox, Angela Snellen, Rebecca DeVries, Sarah Butterfield, Amy Overway, Karen Sweet, Todd Nickoles, Tammy Raterink, Nancy Doyle, Denise Katerberg, Amanda Vandenberg, Cathie Vandersteen, Alesia Derks, Catherine Keezel, Emily Carlson, Kathleen Puff, Andrea Lynn Rosendahl, Kaylee Marcinkus, Barbara Vincensi, Donna Garrett, Jenna Dietsch, and Jody Kinney. Authors have no conflict of interest.
References Anthony, D., Willock, J., & Baharestani, M. (2010). A comparison of Braden Q, Garvin, and Glamorgan risk assessment scales in paediatrics. Journal of Tissue Viability, 19, 98–105. Bergstrom, N., Braden, B. J., Laguzza, A., & Holman, V. (1987). The Braden Scale for predicting pressure sore risk. Nursing Research, 36, 206–210. Bergstrom, N., Demuth, P. J., & Braden, B. J. (1987). A clinical trial of the Braden Scale for predicting pressure sore risk. Nursing Clinics of North America, 22, 417–428. Bernabe, K. Q. (2012). Pressure ulcers in the pediatric patient. Current Opinion in Pediatrics, 24, 352–356. Curley, M., Razmus, I., Roberts, K., & Wypij, D. (2003). Predicting pressure ulcer risk inpediatric patients: The Braden Q Scale. Journal of Eastern Nursing Research Society, 22–33. European Pressure Ulcer Advisory Panel, & National Pressure Ulcer Advisory Panel (2009). Prevention and treatment of pressure ulcers: Quick reference guide. Washington DC: National Pressure Ulcer Advisory Panel. Kottner, J., Wilborn, D., & Dassen, T. (2010). Frequency of pressure ulcers in the paediatric population: A literature review and new empirical data. International Journal of Nursing Studies, 47, 1330–1340. MedCalc Software (2014). MedCalc Easy-to-Use Statistical Software: Diagnostic test calculator (Last modified 05-22-14) [Software]. Available from. http://www.medcalc.org/calc/diagnostic_test.php Noonan, C., Quigley, S., & Curley, M. (2006). Skin integrity in hospitalized infants and children. Journal of Pediatric Nursing, 21, 445–453. Parnham, A. (2012). Pressure ulcer risk assessment and prevention in children. Nursing Children and Young People, 24, 24–29. Pediatric Affinity Group (2014). How-to-guide pediatric supplement: Preventing pressure ulcers (Last updated 02–01–14). Available from. http://www.nichq.org/pdf/FINALPressureUlcers. pdf Polit, D. F., & Beck, C. T. (2013). Essentials of nursing research: Appraising evidence for nursing practice (8th ed.). Philadelphia, PA: Lippincott Williams & Wilkins. Quigley, S., & Curley, M. (1996). Skin integrity in the pediatric population: Preventing and managing pressure ulcers. Journal for Specialists in Pediatric Nursing, 1, 7–18.
610 Sim, J., & Wright, C. C. (2005). The Kappa statistic in reliability studies: Use, interpretation, and sample size requirements. Physical Therapy, 85, 257–268. VanderKooi, M., Blackport, C., & Vander Laan, K. J. (2011). Nurses’ information flow and communication: A comparison of paper to electronic Kardex – Part 3 survey. Poster presented at Spectrum Health Nursing Research Traveling Poster Exhibit, Grand Rapids, Michigan. Waterlow, J. (1997). Pressure sore risk assessment in children. Paediatric Nursing, 9, 21–24.
D.J. Sterken et al. Willock, J., Anthony, D., & Richardson, J. (2008). Inter-rater reliability of the Glamorgan Pediatric pressure ulcer risk assessment scale. Pediatric Nursing, 20, 14–19. Willock, J., Baharestani, M. M., & Anthony, D. (2009). The development of the Glamorgan paediatric pressure ulcer risk assessment scale. Journal of Wound Care, 18, 17–21. Willock, J., Harris, C., Harrison, J., & Poole, C. (2005). Identifying the characteristics of children with pressure ulcers. Nursing Times, 101, 40–43.